David A Olson, MD

Head injury can be defined as any alteration in mental or physical functioning related to a blow to the head. Loss of consciousness does not need to occur. The severity of head injuries is most commonly classified by the initial postresuscitation Glasgow Coma Scale (GCS) score, which generates a numerical summed score for eye, motor, and verbal abilities.
Traditionally, a score of 13-15 indicates mild injury, a score of 9-12 indicates moderate injury, and a score of 8 or less indicates severe injury. In the last few years, however, some studies have included those patients with scores of 13 in the moderate category, while only those patients with scores of 14 or 15 have been included as mild. Concussion and mild head injury are generally synonymous.
Research on head injury has advanced considerably in the last decade. As is typical of many endeavors, these efforts have exposed the complexity of this condition more deeply and have helped researchers and physicians to abandon crude simplifications. This review concentrates primarily on current developments in the diagnosis and management of closed head injuries in adults.
Pathophysiology Structural changes
Gross structural changes in head injury are common and often obvious both on autopsy and conventional neuroimaging. The skull can fracture in a simple linear fashion or in a more complicated depressed manner, in which bone fragments and pushes beneath the calvarial surface. In patients with mild head injury, a skull fracture markedly increases the chance of significant intracranial injury.
Both direct impact and contrecoup injuries, in which the moving brain careens onto the distant skull opposite the point of impact, can result in focal bleeding beneath the calvaria. Such bleeding can result in an intracerebral focal contusion or hemorrhage as well as an extracerebral hemorrhage. Extracerebral hemorrhages are primarily subdural hemorrhages arising from tearing of bridging veins, but epidural hemorrhages from tearing of the middle meningeal artery or the diploic veins are also common. Occasionally, subdural hemorrhages can result from disruption of cortical arteries. This type of subdural hemorrhage is rapidly progressive and can occur after trivial head injury in elderly patients.
One study of CT images from 753 patients with severe head injury from the National Institute of Health Traumatic Coma Data Bank in the United States found evidence of intracranial hemorrhagic lesions in 27%. Traumatic subarachnoid hemorrhage was even more frequent and occurred in 39% of patients. Furthermore, diffuse cerebral edema also was present in 39%. Cerebral edema can be unilateral or diffuse and can occur even in the absence of intracranial bleeding. Severe brain edema probably occurs more commonly in children than in adults.
Neuronal loss is also important. A recent pathological study found that quantitative loss of neurons from the dorsal thalamus correlated with severe disability and vegetative state outcomes in patients with closed head injuries.
Finally, axonal injury increasingly has been recognized as a structural sequela of brain injury. The use of amyloid precursor protein staining has resulted in increased recognition of this form of injury. Using this technique, researchers have readily identified axonal injury in patients with mild head injury. Interestingly, a prominent locus of axonal damage has been the fornices, which are important for memory and cognition. More severe and diffuse axonal injury has been found to correlate with vegetative states and the acute onset of coma following injury.
Neurochemical changes
After traumatic brain injury, the brain is bathed with potentially toxic neurochemicals. Catecholamine surges have been documented in the plasma (higher catecholamine levels correlated with worse clinical outcomes) and in the cerebrospinal fluid (CSF) of patients with head injuries (higher CSF 5-hydroxyindole acetic acid (HIAA), the serotonin metabolite, correlated with worse outcomes). Head injury causes release of free radicals and breakdown of membrane lipids. These lipids fragment into mediators of inflammation. The excitotoxic amino acids (ie, glutamate, aspartate) initiate a cascade of processes culminating in an increase in intraneuronal calcium and cell death. Researchers using a microdialysis technique have correlated high CSF levels of excitotoxic amino acids with poor outcomes in head injury.
Although neuroprotective strategies employing antiexcitotoxic pharmacotherapies were effective in diminishing the effects of experimental brain injuries in laboratory animals, clinical trials in humans generally have been disappointing.These failures have prompted development of more complex models of neuronal injury and cell death. Recently, researchers have demonstrated that although certain types of glutamate antagonists may protect against acute cell death, they potentiate slowly progressive neuronal injury in experimental rodent models. Still others have found that low-dose glutamate administered before brain injury is somehow neuroprotective. Such dose and timing effects are only beginning to be understood.
Prostaglandins, inflammatory mediators produced by membrane lipid breakdown, are also elevated dramatically in the plasma of patients with moderate-to-severe head trauma during the first 2 weeks after injury. Patients with higher prostaglandin levels had significantly worse outcomes than those with more modest elevations. Furthermore, levels of a thromboxane metabolite, a potent vasoconstricting prostaglandin, were elevated disproportionately. Such a process may underlie posttraumatic vasospasm, which has been documented in some, but not all, transcranial Doppler studies of patients with closed head injuries, even in patients without traumatic subarachnoid bleeds.
Recently, an increase in T cells reactive against myelin antigens was found in 10 patients with severe head injuries. Although the sample size was limited, those patients with increased T-cell reactivity had improved outcomes compared with their nonreactive counterparts, and a beneficial autoimmune response was proposed.
In addition to structural and chemical changes, gene expression is altered following closed head injury. Genes involving growth factors, hormones, toxin-binders, apoptosis (programmed cell death), and inflammation have all been implicated in rodent models. For example, in a mouse model of head injury, elevated levels of the transcription factor p53 were found. p53 translocates to the nucleus and initiates apoptosis or programmed cell death. Such a process could account for the delayed neuronal loss seen in head injuries.
Secondary insults
Hypotension and hypoxia cause the most prominent secondary trauma-induced brain insults. Both hypoxia and hypotension had adverse impacts on outcomes of 716 patients with severe head injuries from the Traumatic Coma Data Bank in the United States. Efforts to limit hypoxic injury with in-field intubation have been unsuccessful. Indeed, a multicenter study of 4098 patients with severe traumatic brain injury found that in-field intubation was associated with a dramatic increase in death and poor long-term neurologic outcome, even after controlling for injury severity.
In the Trauma Coma Data Bank study, hypotension was even more significant than hypoxia and, by itself, was associated with a 150% increase in mortality rate. Systemic hypotension is critical because brain perfusion diminishes with lower somatic blood pressures. Brain perfusion (ie, cerebral perfusion pressure) is the difference between the mean arterial pressure and intracranial pressure. The intracranial pressure is increased in head injury by intracranial bleeding, cell death, and secondary hypoxic and ischemic injuries. Accordingly, another recent study reported that death and increased disability outcomes correlated with the durations of both systemic hypotension and elevated intracranial pressures.
Severe anemia is often coexistent with head injuries, but blood transfusions have been recently associated with increased mortality and complications among 1,250 ICU-admitted patients with brain injuries. This relationship held even after controlling for the degree of anemia.
Finally, posttraumatic cerebral infarction occurs in up to 12% of patients with moderate and severe head injuries and is associated with a decreased Glasgow Coma Scale, low blood pressure, and herniation syndromes.
Mortality/Morbidity
According to the CDC, 50,000 individuals die from traumatic brain injuries each year in the United States. Almost twice that number suffers permanent disability.
Race
A study of intentional head injury from Charlotte, North Carolina, found minority status was a major predictor of intentional head injury, even after controlling for other demographic factors. Furthermore, worse clinical outcomes have been described for African American children with moderate-to-severe head injuries compared with their white counterparts.
Sex
Men in the United States are nearly twice more likely to be hospitalized with a brain injury than women. This male predominance is found worldwide.
Age
Approximately half of the patients admitted to a hospital for head injury are aged 24 years or younger.
Causes
Road accidents involving motor vehicle drivers and occupants, cyclists, and pedestrians are the main risk factor for head injuries.
Assaults in economically depressed regions and during wartime are other major risk factors.
Athletic participation, especially football and soccer, is another important cause of these injuries.
Falls cause head injuries in elderly patients and children, occasionally with catastrophic results. The incidence of fall-related traumatic brain injury has been increasing in the United States and in 2005 resulted in 7,946 deaths and 56,423 hospitalizations in the elderly.
Blast injuries from incendiary devices can cause head trauma and primarily occur in soldiers, although even civilian tire explosions have been implicated. While the energy from the blast can directly impact the cranium and be transferred to the brain, some researchers have hypothesized that systemic blood vessels may actually transmit the shock waves. Current clinical studies, however, have failed to identify a unique pattern of neuropsychologic deficits in patients who have incurred such blast injuries.
Anticoagulants and antiplatelet medications, such as aspirin, raise the risk of intracranial bleeding with even trivial head injuries. For example, among elderly patients with head injuries, clopidogrel use has been associated with a 15 times greater mortality compared with patients not taking antithrombotics. Alcohol use raises the risks of incurring a head injury.
Perhaps because it may impede excitotoxicity, alcohol use at the time of injury may decrease the likelihood of a poor outcome.
A newer study of intentional head injuries reported that patients consuming alcohol had higher initial GCS scores. Another study of patients with apparently trivial injuries (patients either were found down or fell from heights <10 ft) found that outcomes were better in patients who were severely intoxicated (blood alcohol levels >200 mg/dL). Methamphetamine use has also been shown to reduce mortality in severe head injury.47 More recently, patients with severe brain injuries and high blood alcohol levels (³ 0.08 mg/L) exhibited a significantly lesser mortality compared with patients with lower levels or the absence of alcohol in their blood. Although the presence of APOE4 alleles is not an established risk factor for head injury, the presence of even one of these alleles increases the risk of a poor outcome.
Patients who are homozygous or heterozygous for the APOE4 allele have an almost 14-times greater likelihood of a poor outcome after head injury than those with other APOE genotypes.
Similarly, football players and boxers with an APOE4 allele are at greater risk for posttraumatic cognitive problems than APOE4 -negative athletes.
Other studies have called these APOE4 associations into question, but a 2008 meta-analysis has supported these observations.
Genes regulating the interleukin, dopamine, and apoptotic systems as well as genes associated with angiotensin converting enzyme and calcium channel polymorphisms have all been implicated in head injury outcomes. Other genetic determinants of head injury will undoubtedly surface with further research.
Treatment
Medical Care
Acute management
In the setting of acute head injury, give priority to the immediate assessment and stabilization of the airway and circulation. Despite the fact that prehospital intubation has become common, at least one study has reported a higher rate of mortality in patients intubated in the field than in those intubated in the hospital setting. In this study, however, more critically ill patients required in-field intubation.
Following stabilization, direct attention to prevention of secondary injury. Keep mean arterial pressures above 90 mm Hg; arterial saturations should be greater than 90%. Urgent CT scanning is a priority.
Next, focus attention on reducing intracranial pressure, since elevated intracranial pressure is an independent predictor of poor outcome.
If the intracranial pressure rises above 20-25 mm Hg, intravenous mannitol, CSF drainage, and hyperventilation can be used. Hypertonic saline has also been used in lieu of mannitol to lower intracranial pressure, but more definitive studies are needed.
If the intracranial pressure does not respond to these conventional treatments, high-dose barbiturate therapy is permissible, despite the fact that no evidence currently suggests that barbiturate treatment actually improves outcomes. (Its blood pressure-lowering effects may be detrimental.)
Interestingly, a 2008 study utilizing the National Trauma Data Bank retrospectively uncovered a 45% reduction in survival in patients who underwent intracranial pressure monitoring. These results have been called into question, however, because of a dearth of clinical and neuroimaging data.
Another approach used by some clinicians is to focus primarily on improving cerebral perfusion pressure as opposed to intracranial pressure in isolation.
One study reported that 80% of patients with severe head injuries experienced recoveries with no or little disability after volume expansion, mannitol, CSF drainage, and vasopressors were used to maintain a cerebral perfusion pressure of at least 70 mm Hg. Other studies have found higher perfusion pressures were associated with more complications and have recommended maintaining a cerebral perfusion pressure of 50-70 mm Hg.
The question whether saline or albumin fluid resuscitation would maximize cerebral perfusion pressure and lead to improve outcomes lead to a large, double-blind, randomized controlled study of 460 patients with Glasgow Coma Scale scores <13 who also had abnormal head CT scan results. A post-hoc 2-year follow-up demonstrated increased mortality in those receiving albumin as opposed to saline.
Head injury induces a hypermetabolic state and early nutritional interventions may be as critical as cerebral perfusion pressure. Parental or enteral feedings reduced mortality by at least 50% in one study when given early in the course of severe head injury.
As mentioned previously, head injury may alter coagulation parameters, and this can raise the risk of deep venous thrombosis to as much as 15% if no pharmacologic prophylaxis is given within the first 48 hours. The risk of extension of intracranial bleeding needs to be balanced with the benefits of thromboembolic prevention.
A retrospective review suggested that early prophylaxis is safe because there was no difference between intracranial hemorrhage progression in patients with head injury who received enoxaparin or heparin within the first 3 days versus later in the course of their hospitalization.8 Further studies, of course, are required.
Steroids have demonstrated no benefit in the treatment of acute head injury. A 2004 multicenter European randomized trial of steroids versus placebo found a higher mortality after only 2 weeks in the steroid-treated patients.
Phenytoin has demonstrated efficacy in controlling early posttraumatic seizures, but mortality rates, surprisingly, were unaffected by this benefit. In one study, approximately 2.5% of patients treated with phenytoin had an allergic reaction to the drug during the first 2 weeks of therapy.
A trial of valproate in early seizure prophylaxis showed a trend toward an increased mortality rate. Anticonvulsant therapy, if used, should be discontinued after 1-2 weeks unless further seizures supervene.
Finally, as stated previously, neuroprotective agents mostly have failed to improve the outcomes of patients with brain injury.
However, the calcium channel blocker nimodipine was successful in reducing rates of death and severe disability when instituted acutely in patients with head injuries and traumatic subarachnoid hemorrhages, despite its failure to improve outcomes in 2 large trials of patients with all types of traumatic intracranial injuries.
Although numerous synthetic neuroprotective agents are under development, several existing substances have shown promise, but other agents have been disappointing.
Because of its excitotoxic blocking properties, magnesium chloride has been used to reduce cortical injury in experimentally brain-injured rats.
Unfortunately, a human double-blind study of 499 patients with moderate or severe head injury failed to show benefit; the magnesium-treated patients actually did worse. One potential confounder in this study was vigilance and aggressive repletion of hypomagnesemia in controls.
Progesterone given intravenously in a phase II, randomized, double-blind, placebo-controlled trial of 100 patients with moderate and severe head injury showed no adverse effects and reduced 30-day mortality by 57%.
Unfortunately, worse outcomes were seen in the treated group with severe head injuries as measured by the extended Glasgow Outcome Score, perhaps because of the increased survivorship of sicker patients.
Experimental brain injury creates permeability in mitochondrial membranes, which contributes to cell death by causing calcium effluxes and energy depletion.
Cyclosporin inhibits mitochondrial permeability and has been used in a phase II study of patients with traumatic brain injuries. Further trials are planned.
Cannabinoids also protect against excitotoxicity, but disappointingly, in a recent phase 3 trial, dexanabinol, a weak N -methyl-D-aspartic acid (NMDA) antagonist, showed no efficacy in outcome improvement when given within 6 hours to patients with severe closed head injuries.
Rosuvastatin given in the acute phase of moderate head injury significantly reduced amnesia in a double-blind placebo-controlled study of 34 patients.
Animal studies of some health food supplements may lead to new directions. The dietary supplement creatine, when fed to rats for 4 weeks prior to an experimental brain injury, reduced cortical damage by 50%, primarily through stabilizing mitochondrial functioning. Melatonin is a free radical scavenger, and when injected early in brain-injured rats, it significantly reduced levels of lipid breakdown products.
Long-term management
Hypertonicity from spasticity or dystonia with attendant muscle spasms is often disabling. Although dantrolene, baclofen, diazepam, and tizanidine are current oral medication approaches to this problem, baclofen and tizanidine are customarily preferred because of their more favorable side effect profiles.
When using these agents, careful evaluation of functional status and symptom relief is a priority since adverse effects such as sedation may be pronounced.
Intrathecal baclofen is a newer approach with reported efficacy and minimal adverse effects. One study of 17 patients with traumatic brain injuries showed improved motor tone and decreased muscle spasms with intrathecal baclofen, but whether these benefits will translate into improved functioning remains unknown.
Botulinum toxin also has shown promise in decreasing hypertonia in patients with head injuries, primarily by improving passive range of motion rather than by decreasing functional disability. Solid data on cognitive enhancing medications for patients with head injury are lacking.
Typically, only small numbers of subjects have been use d and demonstrable functional improvement has been only marginally convincing.
Despite these drawbacks, one double-blind, placebo-controlled study of methylphenidate demonstrated improved motor outcomes and attention in patients with head injuries during active treatment, but only 6 patients completed each 30-day treatment arm.
A 2006 double-blind, placebo-controlled study of 18 patients with closed head injuries treated with a single dose of 20 mg of methylphenidate achieved significant improvement in reaction times on a working memory test, but no other cognitive tasks significantly benefited.
Donepezil treatment significantly improved visual and verbal memory as well as attentional deployment in 18 patients with head injuries of all levels of severity in a 2004 double-blind, placebo-controlled study. Other less rigorous studies have also reported cognitive improvements in donepezil-treated, head-injured patients.
Anecdotal reports exist of dramatic alerting responses to both levodopa and methylphenidate in patients with vegetative or comatose states.
Levodopa treatment has also resulted in improvement in patients with akinesia and rigidity secondary to traumatic substantia nigral damage.
Furthermore, levodopa has even produced qualitative cognitive improvements in a small number of head-injured patients.
Emotional lability and the pathologic laughing and crying associated with pseudobulbar palsy reportedly have responded rapidly and exquisitely to selective serotonin reuptake inhibitors.
Sertraline has shown efficacy in depression in mild head injury.
Treat other possible psychiatric complications of head injury on a patient-by-patient basis, since no extensive pharmacologic trials of this dimension of head injury have been conducted.
Surgical Care
Traditionally, the prompt surgical evacuation of subdural hematomas in less than 4 hours was believed to be a major determinant of an optimal outcome.  Indeed, a recent publication found a delay in surgery for acute subdural hematomas of over 5 hours was associated with increased mortality.
Nevertheless, other recent investigations have emphasized that the extent of the original intracranial injury and the generated intracranial pressures may be more important than the timing of surgery.
For example, 70% of 83 patients with GCS scores of 11-15 who had subdural hematomas less than 1 cm in width and no cisternal effacement on neuroimaging or focal neurological deficits were successfully managed nonoperatively with only 6% eventually requiring surgery.
Another study of 462 patients with head injuries with CT-imaged intracranial hematomas who were treated nonoperatively found that only approximately 10% progressed clinically and eventually required surgery.
Frontal parenchymal hematomas were more likely to require eventual surgery.
Decompressive craniectomies are sometimes advocated for patients with increased intracranial pressure refractory to conventional medical treatment.
Of 40 patients with severe head injury who underwent this procedure (some for ICP elevations in isolation and some for ICP elevations with mass lesions), 30% had a favorable long-term outcome.
 At least 2 major randomized clinical trials of this intervention are now underway. The operative and nonoperative management of intracranial injuries is an ever-evolving area of study and, at present, more a matter of neurosurgical judgment than hard and fast decision rules.
Medication
Medications commonly are used in the acute setting to control early seizures, reduce intracranial pressure, and correct electrolyte abnormalities. Nimodipine may be neuroprotective in the subset of patients with traumatic subarachnoid hemorrhages.
In the long-term setting, cognitive and motoric augmentation as well as the control of spasticity and emotional incontinence may require pharmacologic interventions.

Source: emedicine from WebMD