TRAUMA TO THE CENTRAL NERVOUS SYSTEM (also see Chapter 53 )

CNS trauma accounts for almost half of all trauma deaths examined postmortem in population-based analyses^[117] or after trauma center admission.^[118] In addition, more than 90,000 Americans are disabled by TBI^[119] and 8000 to 10,000 by SCI^[120] each year. As with hemorrhagic shock, CNS trauma consists of both the primary injury, in which tissue is disrupted by mechanical force, and a secondary response, in which the body's reaction to the injury plays an important role. Mitigation of secondary CNS injury depends on prompt diagnosis and early, goal-directed therapy. Although there is no way to minimize primary CNS injury other than by prevention strategies, it is now apparent that secondary brain injury accounts for much of the death and disability after trauma. The initial management of these patients can significantly affect the outcome. Attention to the ABCDE approach is paramount to successful resuscitation, and trauma anesthesiologists should be intimately involved in this process.

Traumatic Brain Injury

Approximately a third of the 150,000 deaths after traumatic injury in the United States each year are due to fatal head injuries.^[121] Of the 1 million survivors of TBI annually, almost 25% require inpatient care. The most severely brain injured have the highest mean length of stay and mean hospital costs. Regardless of the initial severity of injury, TBI patients are at high risk for disabilities, and patients who survive after TBI are more impaired than other trauma cohorts. The public health burden of long-term rehabilitation of patients with severe TBI accounts

2470

Research into treatment modalities to improve outcome from TBI has been disappointing, perhaps because of the diffuse nature of the brain injury and systemic response. Several agents and techniques that have shown promise in the laboratory setting have not proved to be effective in clinical trials. Perhaps the mechanisms of secondary brain injury in humans are not the same as those in animals, or perhaps other systemic injuries or premorbid conditions contribute to the overall poor outcome in trauma patients. Drugs such as free radical scavengers, anti-inflammatory agents, and ion channel blockers that have been effective in animals have had little impact or disappointing results in human trials.^[122]

Brain injury after trauma is classified as mild, moderate, or severe, depending on the GCS score on admission. The GCS was introduced in 1971 as a means of grading traumatic coma but rapidly progressed to a classification of head injury severity.^[9] Mild traumatic brain injury (GCS score of 13 to 15) is the most common and generally requires only cranial CT to assess anatomic injury and rule out operative pathology, followed by a short period of observation. Deterioration of young patients with mild TBI is rare because limited intracranial compliance will lead to depression of mental status early after injury. Elderly patients are more apt to experience a "honeymoon" period of near-normal GCS scores after TBI because the pathophysiologic effects of intracranial hemorrhage and edema are delayed by compensatory relocation of cerebrospinal fluid (CSF) outside the cranium. Patients with mild TBI who maintain a stable GCS score for 24 hours after the injury are very unlikely to deteriorate further, although they are at risk for a number of "postconcussive" effects, including headaches, memory loss, emotional lability, and sleep disturbance.^[123]

Moderate TBI (GCS score of 9 to 12) may be manifested as intracranial lesions that require surgical evacuation, and early cranial CT is strongly indicated. These patients have a higher potential for deterioration than those with mild TBI. Early intubation, mechanical ventilation, and close observation may be required in the management of patients with moderate TBI because of combative or agitated behavior and the potentially catastrophic consequences of respiratory depression or pulmonary aspiration occurring during diagnostic evaluation. Extubation can be undertaken if the patient is hemodynamically stable and appropriately responsive after the diagnostic workup. Treatment of secondary brain injury is accomplished by early correction and subsequent avoidance of hypoxia, prompt fluid resuscitation, and management of associated injuries. The timing of indicated noncranial surgery in these patients is highly controversial inasmuch as early surgery has been associated with an increase in episodes of hypoxia and hypotension.^[124] Although some studies have demonstrated a decrement in neurologic outcomes after early surgery,^{[125] [126]} others have shown an increased incidence of pulmonary and septic complications when orthopedic and soft tissue surgery is delayed.^{[127] [128]} Recent reviews of surgical timing have not substantiated an increased risk associated with early surgery, although a definitive prospective trial has not yet been reported.

Neurologic monitoring of patients with moderate TBI consists of serial assessment of consciousness and motor and sensory function. Deterioration of the GCS is an indication for urgent repeat cranial CT to establish the need for craniotomy or invasive monitoring of ICP. If frequent neurologic monitoring is not possible because of general anesthesia continuing more than 2 hours, the need for aggressive analgesia, or prophylaxis against delirium tremens, invasive ICP monitoring is indicated.^[122] Although mortality from moderate TBI is low, almost all patients will suffer significant long-term morbidity.

Severe TBI is classified as a GCS score of 8 or less at the time of admission and carries a significant risk for mortality. Early, rapid management focused on restoration of systemic homeostasis and perfusion-directed care of the injured brain will produce the best possible outcomes in this difficult population.

Airway and Ventilatory Management (also see Chapter 42 )

A single episode of hypoxemia (PaO₂ <60 mm Hg) occurring in a patient with severe TBI is associated with a near doubling of mortality.^[129] Prehospital intubation is indicated and now occurs routinely in the increasing number of emergency medical systems that have adopted medication-facilitated early intubation protocols. The patient should be transported as rapidly as possible to a facility capable of managing severe TBI or to the nearest facility capable of intubating the patient and initiating systemic resuscitation. Some data suggest that patients with severe TBI who are intubated in the field may have better outcomes^[130] ; however, the sine qua non is adequacy of systemic oxygenation, by whatever means it can best be accomplished.

Patients with isolated head injuries can be managed with traditional ventilatory strategies, but those with chest trauma, aspiration, or massive resuscitation after shock are at high risk for acute lung injury. The classic teaching of no or low-level positive end-expiratory pressure (PEEP) to prevent elevated ICP is inappropriate because it may fail to correct hypoxemia. With adequate volume resuscitation, PEEP does not increase ICP, nor does it lower cerebral perfusion pressure (CPP)^[131] ; however, it may actually decrease ICP because of improved cerebral oxygenation.^[132]

Hyperventilation therapy, long a mainstay in the management of patients with TBI, is no longer an appropriate treatment unless signs of imminent herniation are present. Cerebral blood flow (CBF) decreases after TBI; iatrogenic hyperventilation may further decrease blood flow to ischemic levels and worsen secondary injury.^[133] Although decreased PaCO₂ decreases CBF in brains with intact autoregulation, vasoreactivity is not always intact after TBI.^[134] Therefore, hyperventilation has the possibility of inducing hyperemia in areas of the brain that do not have the intact autoregulatory mechanisms that occur in other regions of the brain as a result of vasoconstriction. Because it is impossible to predict which patients have intact autoregulation and vasoreactivity based on the type of head injury, institution of this therapy purely

2471

Circulation

The most challenging of all trauma patients are those with severe TBI and coexisting hemorrhagic shock. A single episode of hypotension, defined as systolic BP less than 90 mm Hg, is associated with an increase in morbidity and doubled mortality after severe TBI.^[129] Hypotension together with hypoxia is associated with threefold mortality. Patients with severe TBI should have systolic BP higher than 110 mm Hg with a goal of achieving MAP greater than 90 mm Hg (to allow for a minimum CPP of 70 mm Hg) until ICP monitoring is instituted and CPP can be directly targeted.

The American Association for Neurological Surgeons (AANS) and the Brain Trauma Foundation guidelines for the management of patients with severe TBI suggest maintenance of CPP at a minimum of 70 mm Hg at all times. Attempts to increase perfusion to greater levels have not generally shown a corollary improvement in outcome. Studies have demonstrated that there is no direct relationship between CBF and ICP after TBI; this finding is true whether autoregulation is intact or is lost.^[134] Contrary to practices in the past, current recommendations are to maintain patients with severe TBI in a euvolemic state. Therefore, fluid resuscitation is the mainstay of therapy, followed by vasoactive infusions as needed. The ideal fluid has not yet been identified, but increasing evidence suggests that hypertonic saline solutions are optimal. Correction of anemia from blood loss is the first priority, with a goal of maintaining hematocrit greater than 30%. After the initial ABCDE management of a patient with severe TBI, a stepwise approach to maintenance of CPP is initiated.

Management of Intracranial and Cerebral Perfusion Pressure

The critical threshold for ICP management has not been clearly defined. Most authors support treatment of ICP greater than 20 to 25 mm Hg, but higher values may be tolerated if CPP is adequate. Unfortunately, patients can herniate at pressures lower than 20 mm Hg, depending on the location of their lesion, so each case must be treated individually.^[135] Communication between the neurosurgeon and the anesthesiologist/intensivist is critical.

According to AANS guidelines, there are several indications for ICP monitoring after TBI.^[122] Patients with severe head injury (defined as a GCS score less than 8) and abnormal head CT findings (hematoma, contusions, edema, or compressed basal cisterns) should be managed with the aid of ICP monitoring. In addition, patients should be monitored if they have severe TBI, normal head CT findings, and any of the following: age older than 40 years, motor posturing, or systolic BP lower than 90 mm Hg. A high correlation has been found between severe TBI, elevated ICP (>20 mm Hg), and poor outcome. Because aggressive management of CPP is correlated with improved outcome and the placement of an ICP monitor is considered to be a low-risk procedure in most patients, it should be used in the aforementioned situations unless the risk of inducing intracranial hemorrhage is high. Nonetheless, there have been several reports in the recent literature of poor compliance with these recommendations.^[136]

Mortality in patients with severe TBI and elevated ICP is decreased when intraventricular drainage (ventriculostomy) is used as opposed to intraparenchymal monitoring (fiberoptic filament or extradural catheter) to guide ICP/CPP management. ^[137] Current AANS recommendations state that ventriculostomy is the most accurate, safe, and cost-effective method to monitor ICP.^[122] In addition to its diagnostic capabilities, an intraventricular catheter monitor allows therapeutic treatment by drainage of CSF. Complications such as infection and hemorrhage are relatively infrequent and should not prevent placement when clinically indicated.

Because CPP values do not assess a physiologic end point, the ability to determine tissue oxygenation is a valuable measure in the management of patients with severe TBI. Current AANS guidelines do not make recommendations on the use of these technologies, but many institutions routinely use monitors of cerebral tissue O₂ utilization. Jugular venous oxygen saturation represents venous drainage in the jugular bulb and may be used to guide therapy. Desaturation of less than 55% is associated with a worse outcome,^[138] but patients with elevated jugular venous O₂ saturation (>75%) may also experience a lower cerebral metabolic rate of oxygen (CMRO₂ ),^[139] and no direct correlation has been found between this monitoring technique and CBF. Noninvasive measurement of CBF with acoustic technology has been reported but is not yet validated for clinical use.^[140] Near-infrared spectroscopy for determining cerebral oxygen saturation after head injury has not been shown to consistently assess changes caused by CO₂ or arterial pressure changes,^[141] and brain tissue oxygen saturation techniques have not as yet demonstrated a definitive outcome benefit.^[142] The advent of cerebral microdialysis permits sampling of extracellular metabolites and neurotransmitters that are reflective of cerebral metabolism, and though still primarily a research tool, microdialysis is now in clinical use in some centers. ^[143] Delivery of oxygen to the brain remains the primary goal in treating severe TBI, but the best method for assessing oxygen delivery has not yet been established.

Once monitoring of the brain is established, therapy is initiated according to the stepwise recommendations of the Brain Trauma Foundation. Figure 63-11 is a graphic depiction of this approach; the individual therapies are discussed in the following paragraphs.

Positional therapy is a mainstay of treatment of severe TBI. Elevation of the patient's head will facilitate venous drainage and movement of CSF from the cranium to the spinal canal, thereby lowering ICP. Although a reduction in CPP will also occur, the effect on ICP will outweigh this reduction as long as the patient is euvolemic. CPP will thus improve when the patient is in a sitting position.

2472

Figure 63-11 Escalating therapy for severe traumatic brain injury. The goal of therapy is to maintain cerebral perfusion pressure above 70 mm Hg by support of the circulation and control of intracranial pressure. Progressively more intensive therapies are added until this goal is achieved. Decompressive craniotomy is controversial, and decompressive laparotomy is under study at this time. CSF, cerebrospinal fluid.

Analgesics are indicated for the treatment of pain arising from coexisting injuries because catecholamine secretion in response to pain may increase CMRO₂ . The use of sedatives after TBI should be judicious and tempered by the fact that their use may impede sequential neurologic assessment. With this understanding, it is important to recognize that sedatives are a valuable and often necessary adjunct for control of elevated ICP. Short-acting agents are preferred, and propofol is desirable because of this quality. In addition, patients with elevated ICP who are sedated with propofol (in addition to a narcotic) have been shown to maintain better control of ICP and require less neuromuscular blockade and additional CNS-depressant therapy.^[144] Propofol does not appear to have direct vasoregulatory action on cerebral vasculature, but it is a negative inotrope and systemic vasodilator that may decrease MAP—and thus CPP—if not administered carefully.^[145] The use of propofol to decrease ICP frequently mandates the use of vasoactive drugs to maintain MAP. Though appropriate under these conditions, the intensivist must be wary of systemic ischemia arising from the use of pressor drugs; invasive hemodynamic monitoring with a pulmonary artery catheter and frequent assessment of lactate and base deficits may be necessary to maintain an appropriate intravascular volume in the face of confounding pharmacologic drugs. Neuromuscular blocking drugs may be necessary for temporary control of severely elevated ICP. Because neuromuscular blockers interfere with the neurologic examination and can result in significant side effects after prolonged use,^[122] their use should be limited to intraoperative administration and as a temporizing measure while other therapies to control ICP are begun.

Mannitol has long been used as first-line therapy to control acute elevations in ICP. Although it effectively reduces ICP through its actions as an osmotic diuretic and cerebral arteriolar vasoconstrictor, it has been associated with rebound intracranial hypertension.^[146] This hypertension may be caused by intravascular dehydration and an inability to maintain CPP when diuresis has not been matched by appropriate fluid administration. Mannitol is contraindicated in a patient who is not adequately volume resuscitated, and every effort should be made to maintain euvolemia in patients throughout the treatment of elevated ICP. Recent studies investigating the use of hypertonic saline solutions have been promising. The beneficial effects of HS are due to several mechanisms. In addition to its osmotic effect on edematous cerebral tissue, HS also exerts hemodynamic, vasoregulatory, immunologic, and neurochemical effects.^[147] Increases in BP and cardiac output not only are due to plasma volume expansion but may also occur because of changes in circulating hormone levels.^[148] Vasospasm occurring after TBI may be counteracted by HS through vasodilatory actions.^[149] The perturbations in extracellular sodium and excitatory neurotransmitters that occur after injury may be attenuated by HS, and its depression of leukocyte adherence and neutrophil margination may offer protection from bacterial illnesses.^[150] Concentrations of HS used in clinical trials range from 1.7% to 29.2%, and it consistently produces significant decreases in ICP, improvements in CPP, and often, enhanced hemodynamics. Despite the theoretical concern for hypernatremia, it has not been found to be clinically detrimental, and concerns of hyperchloremia can be addressed by the addition of acetate to the solution to buffer the effects of acidosis.

A small percentage of patients will manifest intractable ICP elevations that may respond to barbiturate coma. In addition to lowering CMRO₂ , barbiturates have been shown to result in a decrease in excitatory neurotransmitters (glutamate, lactate, aspartate) by microdialysis analysis.^[151] Barbiturate therapy is considered only when CPP cannot be maintained by the previously described therapies. Management of barbiturate coma necessitates exquisite management of intravascular volume—almost certainly requiring pulmonary artery catheterization—and frequent use of vasoactive agents.

Decompressive craniectomy is a surgical procedure used to control severely elevated ICP and prevent herniation after stroke, and it is now more widely used for the same indications after severe TBI. Decompressive craniectomy is indicated for selected anatomic patterns of TBI when CPP cannot be maintained despite vigorous application of the previously described therapies, including barbiturate coma. Recent evidence suggests that relieving the increased ICP by

2473

Like hyperventilation therapy, there has been a progressive change in the use of hypothermia to treat severe TBI. Early studies demonstrated that moderate, systemic hypothermia reduced the rate of both cerebral edema and mortality after cortical injury in laboratory animals.^{[153] [154]} Small clinical series in humans also suggested improved outcome in patients suffering TBI when hypothermia was maintained for 24 or 48 hours.^{[155] [156]} However, a recently published, multicenter randomized trial of hypothermia (33°C) versus normothermia demonstrated no improvement in outcome in a population of patients with severe TBI.^[157] Of note, patients who were hypothermic on admission and then randomized to the normothermia group had a worse outcome than did those who were left hypothermic, thus leading to the recommendation that patients with severe TBI who are hypothermic on admission not undergo active rewarming. Hyperthermia increases CMRO₂ and has been shown to worsen findings on neurologic examination.^[158] Though not commonly seen on admission, fever may develop in critically ill patients with TBI during their ICU stay. Aggressive attempts to return the temperature to normal should be instituted, with special care not to induce shivering, which also increases CMRO₂ .

Management of Intracranial Pressure in the Operating Room (also see Chapter 53 )

Although most management of patients with severe TBI will occur in the ICU, urgent cranial or noncranial surgery is frequently indicated. Because of the risks of transport to and from the OR and the potential hemodynamic perturbations that can occur during surgical and anesthetic care, operations on patients with severe TBI should be kept to a minimum. Control of life- or limb-threatening pathology is indicated at any time, whereas urgent but not emergency procedures such as fixation of long bone fractures should be undertaken only if CPP is being successfully managed. All of the previously described therapies should be continued throughout the perioperative period, including positional therapy (when possible), aggressive hemodynamic monitoring and resuscitation, administration of osmotic agents (with attention to maintaining euvolemia), and deep levels of analgesia and sedation. Appropriate anesthetic choices include narcotics and low concentrations of volatile anesthetics. Benzodiazepine administration may interfere with subsequent neurologic assessment, and nitrous oxide is contraindicated because of laboratory evidence suggesting less favorable outcomes after its use in animal models of TBI.^[159]