Pathophysiology and Clinical Management of Moderate and Severe Traumatic Brain Injury in the ICU

Abstract

Moderate and severe traumatic brain injury (TBI) is the leading cause of morbidity and mortality among young individuals in high-income countries. Its pathophysiology is divided into two major phases: the initial neuronal injury (or primary injury) followed by secondary insults (secondary injury). Multimodality monitoring now offers neurointensivists the ability to monitor multiple physiologic parameters that act as surrogates of brain ischemia and hypoxia, the major driving forces behind secondary brain injury. The heterogeneity of the pathophysiology of TBI makes it necessary to take into consideration these interacting physiologic factors when recommending for or against any therapies; it may also account for the failure of all the neuroprotective therapies studied so far. In this review, the authors focus on neuroclinicians and neurointensivists, and discuss the developments in therapeutic strategies aimed at optimizing intracranial pressure and cerebral perfusion pressure, and minimizing cerebral hypoxia. The management of moderate to severe TBI in the intensive care unit is moving away from a pure “threshold-based” treatment approach toward consideration of patient-specific characteristics, including the state of cerebral autoregulation. The authors also include a concise discussion on the management of medical and neurologic complications peculiar to TBI as well as an overview of prognostication.

Few diseases impact global heath to the extent that traumatic brain injury (TBI) does. It is the leading cause of morbidity and mortality among young individuals in high-income countries.¹ Close to 52,000 patients with TBI die each year,² and moderate to severe TBI disables 80,000 persons per year.³

Classification

Traumatic brain injury may be classified by etiology or severity. Classification by etiology includes blunt closed head injuries, penetrating head injuries, and blast head injuries. Traumatic brain injury may also be classified by severity into mild, moderate, and severe. A Glasgow Coma Score (GCS) of 13 to 15 is considered mild, 9 to 12 moderate, and GCS of 8 or less severe.⁴ An alternative method of classification divides TBI into focal and diffuse. Focal injury constitutes extra-axial hematomas (subdural and epidural), intraparenchymal hematomas, and hemorrhagic contusions.⁵ Diffuse brain injury involves widespread damage to axons, diffuse micro\vascular injury, hypoxic ischemic injury, and brain swelling.⁵ Axonal injury (also referred to as diffuse axonal injury) is the result of shear, tensile, and compressive forces within brain tissue due to heterogeneity in tissue consistency and degree of fixation. It usually occurs in the setting of high-velocity motor vehicle accidents.

Pathophysiology

Traumatic brain injury may be divided into two major phases: the initial neuronal injury (or primary injury) followed by secondary insults (secondary injury).³ The primary injury is the initial neuronal injury that occurs immediately and is a direct result of the traumatic event. The primary injury may not always be treatable and thus often proves fatal. On the macroscopic level, damage includes shearing of white matter tracts, focal contusions, hematomas, and diffuse edema. At the cellular level, early sequelae of neurotrauma include microporation of membranes, leaky ion channels, and stearic conformational changes in proteins. Blood vessels can be torn at higher shear rates, causing microhemorrhages.¹ It cannot be overemphasized that the mainstay of minimizing damage due to primary injury is through prevention.

The secondary injury develops over hours to days, and includes neurotransmitter release (excitotoxicity), free radical generation, calcium-mediated damage, gene activation, mitochondrial dysfunction, mass affect, ischemia, and inflammatory responses.¹

The pathophysiology of TBI both focal and diffuse is multifaceted and heterogeneous, a fact that must be taken into account when designing neuroprotective treatments. This fact might partly explain why neuroprotective therapies to date have not shown significant clinical benefit in humans.

Diagnosis

Traumatic brain injury is a clinical diagnosis, relying on symptoms such as documented loss of consciousness, confusion, or amnesia (anterograde or retrograde) after an impact to the head.¹ The preliminary neurologic exam at the scene should focus on the Glasgow Coma Scale (GCS) and pupillary exam.^6,7 Neuroimaging should be performed as soon as hemodynamic stabilization is achieved, but should be expedited if the patient is comatose or has a focal neurologic deficit.⁶ Magnetic resonance imaging (MRI) is seldom useful in the acute phase of TBI. However, in the diagnosis of diffuse axonal injury (DAI), MRI offers better detection of white matter lesions.⁸ Magnetic resonance imaging also has specific prognostic applications that will be discussed later in this review.

Intensive Care Unit Monitoring

Given the fact that secondary brain injury is an ongoing process exacerbated by inadequate cerebral perfusion and oxygenation, the focus of care in the intensive care unit (ICU) is on monitoring parameters, which act as surrogates of brain ischemia and hypoxia. In the following sections, we will address these parameters in greater detail.

Neurologic Examination

The neurologic exam is the cornerstone of physiologic monitoring of TBI patients in the ICU.⁹ Sequential clinical evaluation is crucial. The most important aspects are assessment of consciousness, usually performed by using the GCS, with a score of 8 or less defining coma. The full outline of unresponsiveness (FOUR) score includes assessment of pupillary light and corneal reflexes, breathing pattern examination, and motor and eye exams; it may be superior to the GCS in predicting in-hospital mortality after TBI.^9–11 Despite the advent of multimodality monitoring (MMM), the neurologic exam remains indispensable in the ICU monitoring of TBI patients and complements other modalities.

Intracranial Pressure and Cerebral Perfusion Pressure

Measuring intracranial pressure (ICP) and cerebral perfusion pressure (CPP) are also important. Conventionally, ICP greater than 20 mm Hg is considered intracranial hypertension (ICH), and was associated with worse outcomes in the Traumatic Coma Data Bank (TCDB) upon retrospective analysis.^12,13 The principle aims of ICP monitoring are to enable early detection of secondary hemorrhage and to guide therapies that minimize ICH.¹³ Also important is the use of ICP to calculate CPP, using the following equation: CPP = MAP − ICP.

In light of the BEST-TRIP trial,¹⁴ there is now greater understanding that treatment based solely on ICP is insufficient to influence TBI outcomes. Although one singular value (20 mm Hg) as a threshold for intervention is likely not correct for all patients, ICP monitoring should not be abandoned. Tailoring treatment to specific patient characteristics is likely a superior strategy. Additionally, treatment of ICP when a numerical threshold is reached may be an oversimplification because cerebral hypoxia and metabolic dysfunction may still occur with normal ICPs. Instead, ICP may be better managed when considered in the context of parameters such as compliance, ICP waveform morphology, cerebral autoregulation, and ICP impact on cerebral blood flow (CBF) and brain metabolism, and vice versa.^9,15 The Brain Trauma Foundation (BTF) outlined the indications for ICP monitoring in their guidelines published in 2007.¹⁶ These indications are as follows:

• Severe TBI (GCS 3–8) after resuscitation and an abnormal computed tomography (CT) scan: the later included CT scans that revealed hematomas, contusions, swelling, herniation, or compressed basal cisterns.^17,18

• Severe TBI and normal CT if two or more of the following exist: age ≥ 40 years, unilateral or bilateral motor posturing, or systolic blood pressure < 90 mm Hg.¹⁸

• Decompressive craniectomy and evacuation of acute supratentorial intracranial hematomas to help guide further management may be considered; however, these neurosurgical interventions are not included in guidelines.¹⁹

Intracranial pressure is best monitored invasively with a ventricular catheter or an intraparenchymal monitor.^9,16 Invasive techniques, in particular external ventricular drains (EVDs), offer the additional advantage of being able to drain CSF as a therapeutic measure to control elevated ICP. Cerebral perfusion pressure is among the most important factors in the preservation of cerebral blood flow (CBF), and is the stimulus to which the autoregulatory response of cerebral vasculature occurs.²⁰ Treatment aimed at CPP (in addition to ICP) is therefore superior to traditional techniques that focused on ICP alone.²⁰ Achieving optimal CPP is associated with favorable outcome after TBI. Optimal CPP was calculated using the pressure-reactivity index, which is a marker of autoregulation.²¹ This is additional evidence against a pure “threshold-based” treatment approach to ICP/CPP without considering other parameters.

Cerebral Blood Flow

Cerebral blood flow (CBF) can be measured by different techniques. These include transcranial Doppler ultrasonography (TCD), thermal diffusion flowmetry, and laser Doppler flowmetry. Although TCD is the most commonly used technique, it is operator-dependent and may have technical limitations depending on the size of the temporal bone window.

Brain tissue oxygenation is another parameter to be followed. There are four general methods to measure brain oxygenation: jugular venous bulb oximetry, direct brain tissue oxygen tension measurement (PbtO2)_, near-infrared spectroscopy (NIRS), and oxygen-15 positron emission tomography (PET).⁹ Cerebral hypoxia (defined as brain tissue oxygen [PbtO₂] < 10 mm Hg) is associated with worse out-comes after severe TBI.²² However, though some studies show that combined ICP/CPP and PbtO₂-based therapy is associated with better outcomes after severe TBI than ICP/CPP-based therapy alone,^23,24 other studies demonstrated no significant difference.²⁵ BOOST-2 is a randomized, controlled trial aimed at determining whether interventions directed at PbtO₂ result in improved outcomes after severe TBI. Its results are expected imminently. Another controversial issue is which side the probe should be inserted on. Placement of the device on the side of the lesion (injured hemisphere) yields quite different results compared with placement on the contralateral (noninjured) hemisphere, and needs to be studied further.²⁶ Near-infrared spectroscopy is a promising newer noninvasive technique that may have additional uses, such as in the assessment of cerebral autoregulation.⁹ Jugular venous bulb oximetry measures SjvO₂ (oxygen saturation at the jugular bulb), allowing assessment of global brain oxygenation as opposed to more focal or regional assessment with PBtO₂ probes.

Cerebral Metabolism

Cerebral metabolism can be assessed at bedside using cerebral microdialysis, which measures extracellular brain chemistry with an intraparenchymal probe. Importantly, changes in these analytes can precede changes in other physiological variables such as ICP.^27,28 Lactate and pyruvate concentrations provide information about the relative contributions of aerobic and anaerobic metabolism.⁹ Cerebral microdialysis is largely used for research purposes at present, but can be useful for prognostication.

Electrophysiology

Nonconvulsive seizures may occur in up to 40% of severely brain-injured patients, and contribute to worsening secondary brain injury.²⁹ Various epileptiform discharges are seen in patients with TBI and may precede actual seizures.³⁰ Quantitative electroencephalography (EEG) using fast Fourier transformation, in which raw EEG signal is converted into a digital form, has made continuous EEG monitoring practical and its interpretation easier in the ICU.

Biomarkers

Several biomarkers have been examined in TBI; however, the ideal brain marker does not yet exist. The existing biomarkers can be categorized according to their source as primarily neuronal, astroglial, or microglial. Of special importance is the role of genetic factors, in particular the ApoE4 allele, which has been linked in several studies with poorer neurologic outcomes, including death, after TBI.³¹

Treatment

As discussed earlier, the main goal of ICU therapy of moderate to severe TBI is the prevention of secondary brain injury by maintaining adequate cerebral perfusion, which involves targeting ICP/CPP as well avoiding cerebral hypoxia. In addition, optimizing medical management and preventing systemic complications in the ICU plays a crucial role in improving outcomes.

Therapies to Control Intracranial Pressure

The target value for ICP is < 20 mm Hg, and for CPP is 50 to 70 mm Hg.⁶ However, as discussed earlier, TBI management in the ICU is moving away from a pure “threshold-based” treatment approach and relies more on a proper understanding of the Monroe Kelly doctrine and the contribution of the different compartments in the cranium to ICP. These compartments include venous blood, arterial blood, CSF, brain parenchyma, and mass lesions; they are discussed below.

Venous blood can be allowed to drain more effectively by elevating the head of the bed to 30 to 45 degrees and ensuring the neck is straight. Other measures include avoiding insertion of internal jugular catheters, especially during an ICP crisis because placement requires laying the patient supine at 180 degrees, which may impair venous drainage and venous obstruction; thus, congestion may occur with catheters in the jugular position.

Reducing intracranial arterial blood is another strategy to reduce ICP. This can be done by techniques that include hyperventilation, optimizing MAP (target 60 mm Hg) and decreasing metabolism by sedation, and in refractory cases, by cooling. Hyperventilation (HV) causes hypocarbia and vasoconstriction, which reduces CBF and subsequently CBV. However, aggressive HV (PaCO₂ of < 28 mm Hg) can cause severe vasoconstriction resulting in decreases in CBF to ischemic levels. For acute ICP management, clinicians should aim for a PaCO₂ goal of 34 to 36 mm Hg. If ineffective, hyperventilation may be titrated to a goal of 28 to 32 (moderate HV). For recalcitrant ICP, the PaCO₂ goal can be advanced to 25 to 30 mm Hg usually with concomitant jugular venous bulb oximetry or PbtO₂ monitoring. If invasive monitoring is not feasible, continuous monitoring of endtidal CO₂ is essential to ensure regulated hypocarbia.⁶ Hyperventilation should be tapered as soon as possible and prolonged HV should be avoided, particularly in the first 24 hours. Prolonged HV has been shown to worsen outcomes.^16,32 Given the association between CBF and cerebral metabolism (flow-metabolism coupling), the use of sedation reduces ICP by reducing cerebral metabolism.

Pharmacological coma may be considered for patients in whom the ICP remains refractory to maximal medical intervention. Traditionally, this is achieved using high-dose barbiturates with the aim of achieving an EEG burst-suppression pattern; propofol infusion is another reasonable alternative. Regardless of the agent chosen, patient selection is crucial and should include patients who have failed other ICP therapies, have intact autoregulation, are hemodynamically stable (have adequate cardiac reserve), and do not have diffuse injury.^33,34

Therapeutic hypothermia is another measure for reducing ICP refractory to first-line therapies. Mild hypothermia to 35° C for between 48 hours to 5 days effectively reduces ICP, and improves outcomes in such patients.^35,36 Hypothermia may cause shivering, which raises the metabolic rate and should be avoided. Control of the rewarming phase is crucial to avoid rebound ICP and temperature overshooting and should not exceed 0.25°C per hour.³⁷

Removal of CSF may also reduce ICP, particularly in the setting of hydrocephalus. An external ventricular drain (EVD) has the advantage of being used for ICP measurement once connected to an external transducer set-up.

Brain edema reduction in cerebral white matter is best managed by osmotherapy. The most commonly used agents are mannitol and hypertonic saline. Mannitol is to be used in the setting of raised ICP or herniation because prophylactic therapy is of little benefit and is associated with adverse effects such as hypotension and hypovolemia.⁶ Hypertonic saline is the other commonly used osmolar agent. Its effects typically last for hours; hence, it provides more sustained ICP control than mannitol.

The use of steroids is not recommended for ICP management in patients with TBI, even though steroids are commonly used in treating increased ICP due to vasogenic edema from mass lesions in non-TBI patients.¹⁶

Decompressive craniectomy (DC) is another measure that was formerly more often employed in the management of raised ICP in TBI. It is still performed, but less frequently and as an adjunct to evacuation of mass lesions (usually hematomas) for management of malignant cerebral edema and intracranial hypertension (ICH).³⁸ However, DC performed for management of ICH alone is more controversial in light of the DECRA trial (Decompressive Craniectomy in Diffuse TBI), which concluded that while early bifrontotemporal-parietal decompressive craniectomy decreased ICP and length of ICU stay in severe diffuse TBI, it was associated with more unfavorable outcomes.³⁹ The results of the RESCUE ICP trial (Randomized Evaluation of Surgery with Craniectomy for Uncontrollable Elevation of ICP), which is currently nearing completion, are eagerly awaited.

Therapies to Optimize Cerebral Perfusion Pressure

Optimizing CPP can be achieved by decreasing ICP and/or increasing the MAP (see Equation 1). Intravenous fluids should be administered judiciously, aiming for a systolic blood pressure of at least 90 mm Hg per the Brain Trauma Foundation guidelines.⁴⁰ Aggressive fluid resuscitation and use of hypotonic or dextrose containing solutions is counterproductive in TBI. In addition the use of colloids may actually worsen outcomes.⁴¹ Vasopressors are used if the CPP fails to respond to IV fluids resuscitation. Noradrenaline has a more predictable and efficient augmentation of CPP and CBF as estimated by transcranial Doppler,⁴² and improves brain tissue oxygenation as compared with dopamine in TBI patients.⁴³

Therapies to Increase Cerebral Oxygenation

The role of lung protective ventilation strategies with lower tidal volumes and lower PaO₂/FiO₂ ratios cannot be over-emphasized in the prevention of acute lung injury (ALI),⁴⁴ which in turn is an independent risk factor for compromised PbtO₂ in TBI patients.

Management of Complications and Other Neurocritical Care Issues

Fever

Both the degree and duration of early posttrauma hyperthermia are closely correlated with the outcomes of acute TBI.⁴⁵ However, conclusive evidence linking prevention of hyperthermia with improved outcomes in TBI patients is lacking.⁴⁶ In addition, the role of early hypothermia as a neuroprotective strategy could not be demonstrated in a randomized trial.⁴⁷ This may in part be due to the suboptimal effectiveness of existing cooling methods; newer cooling devices may allow this issue to be readdressed. Regardless, strict fever control is recommended despite the lack of definitive evidence linking it with improved outcomes. Of note, therapeutic hypothermia for refractory ICP management is relatively well established.³⁵

Seizures

The incidence of posttraumatic seizures (PTS) in the first 2 years after severe TBI ranges from 21%⁴⁸ to almost 50% after penetrating war-related TBI.⁴⁹ Posttraumatic seizures are classified as either early or late; those occurring before 7 days are early PTS. Seizures exacerbate secondary brain injury by increasing ICP, worsening cerebral hypoxia, and increasing metabolic demand. Nonconvulsive electrographic seizures also occur, and can lead to cerebral metabolic crisis, a delayed increase in ICP, and worse clinical outcomes.²⁹ Current evidence supports the use of antiepileptic drugs (AEDs) to prevent early PTS, but not late PTS.^48,50,51 Risk factors for developing early PTS include penetrating TBI, depressed skull fractures, cortical contusions, and subdural and epidural hematomas. At risk patients should be managed with AEDs for seizure prophylaxis for 7 days.⁵¹ Phenytoin, carbamazepine, and valproate have all been studied for PTS, but none of these reduced the incidence of late posttraumatic seizures. ⁵⁰ Although there is no data to clearly establish the efficacy and tolerability of intravenous levetiracetam over phenytoin, the former is emerging as a favorable alternative in cases where there is risk of significant drug-drug interaction and drug toxicity; however, further studies are needed.⁵² Age, development of early PTS, and EEG findings may assist in predicting the risk of late PTS.⁵³

Coagulopathy

One out of three patients suffering from TBI displays signs of a coagulopathy.⁵⁴ Functional assays such as viscoelastic tests may better phenotype the coagulopathy of TBI and guide targeted therapy.⁵⁵ The prophylactic use of FFP in severe TBI patients was associated with adverse effects, including delayed intracranial hematomas and an increase in mortality.⁵⁶ Although blood and blood product transfusions may be necessary in patients with active bleeding, they carry with them certain risks that need to be carefully weighed against benefits, especially if the hematologic derangements are only mild to moderate.⁵⁷

Hyperglycemia

Deranged glucose metabolism is both a complication of severe TBI as well as an independent marker associated with increased mortality and long-term poor outcomes.⁵⁸ However, correcting these levels with early and aggressive insulin therapy does not necessarily translate into better outcomes.^58,59 In addition, strict glycemic control (< 110 mg/dL) is associated with increased episodes of cerebral hypoglycemia as determined by cerebral microdialysis.⁶⁰ Therefore, although reasonably tight glucose control is still recommended in moderate to severe TBI, care must be taken to avoid hypoglycemic episodes, which are potentially quite injurious to the brain even in the nontraumatic state.

Nutrition

After severe TBI, patients experience a systemic and cerebral hypermetabolic state. Patients given early nutrition have a lower mortality rate and significantly lower rates of infection. Current guidelines recommend nutrition at 140% of resting requirements in nonparalyzed patients, and at 100% in paralyzed patients.⁶

Endocrine Derangements

Traumatic brain injury can be complicated by neuroendocrine abnormalities, most commonly hypopituitarism, which was detected in 21% of patients in one study.⁶¹ This can manifest as increased vasopressor requirements or cerebral salt wasting. In the acute phase, adrenal glucocorticoid insufficiency is managed with replacement-dose hydrocortisone. Diabetes insipidus should be managed aggressively to prevent severe hypernatremia, but once present, the serum sodium should be lowered gradually to avoid malignant rebound cerebral edema.⁶ Recently, neuroendocrine immune dysfunction (involving cortisol dysregulation) has been described in TBI, and is associated with poor outcomes.⁶²

Posttraumatic Vasospasm

Posttraumatic vasospasm (PTV) has an estimated incidence of between 10 to 15%, and typically develops between 12 hours to 5 days.⁶³ Risk factors for the development of PTV include parenchymal contusions and fever.⁶⁴ Vasospasm can be screened for using TCD, CT angiography, or digital subtraction angiography. Posttraumatic vasospasm is particularly a problem after blast TBI, and is associated with the presence of pseudoaneurysms. Therapy is not well established, but open surgical treatment was associated with neurologic improvement, and microballoon angioplasty significantly lowered middle cerebral artery and basilar blood flow velocities in one study.⁶⁵

Paroxysmal Sympathetic Hyperactivity

Paroxysmal sympathetic hyperactivity (PSH) is defined as “a syndrome, recognized in a subgroup of survivors of severe acquired brain injury, of simultaneous, paroxysmal transient increases in sympathetic [elevated heart rate, blood pressure, respiratory rate, temperature, sweating] and motor [posturing] activity.”⁶⁶ Its incidence varies between 8 to 33% in different series.^67,68 Overresponsiveness to afferent stimuli is a recently described hallmark of PSH, and may assist the clinical diagnosis.⁶⁹ Young age and diffuse axonal injury are risk factors for developing PSH.⁷⁰ There are no specific guidelines for managing PSH. The evidence for drug therapy is limited to case reports or case series, and includes opiates, GABA agonists, α- and β-blockers, and dopamine agonists with varying effectiveness.

Prognosis

The most important predictors of outcome after severe TBI include GCS after full resuscitation, age, pupillary reactivity, CT findings, and the presence of major extracranial injury.⁷¹ Though the GCS has been traditionally used for prognostication and has a 70% predictive value,⁶ it has certain limitations that can be addressed using the full outline of unresponsiveness (FOUR) score. The latter was shown to be superior in predicting in-hospital mortality in TBI patients.¹¹ Chemical biomarkers, though used in research, are not routinely applied in clinical prognostication. Neuroimaging is an important tool in prognostication after TBI. Computed tomography findings associated with poor outcomes include compressed or absent basal cisterns, traumatic SAH, midline shift, and intracranial lesions.⁷² Magnetic resonance imaging has added significantly to the accuracy of long-term outcome prediction, especially by the use of quantitative diffusion tensor imaging (DTI) for white matter assessment⁷³; this, however, is not widely available for clinical use yet. Various outcome prediction models have been developed; two such models, referred to as the IMPACT and CRASH models, have been reciprocally validated and were shown to provide adequate discrimination between patients with good and poor 6-month outcomes after TBI using baseline characteristics, especially if CT and laboratory findings (glucose, platelets, hemoglobin) were considered in addition to traditional predictors.⁷⁴ The largest amount of prognostic information was contained in a core set of three predictors: age, motor score, and pupillary reactivity at admission.⁷⁴ There are no guidelines as to the optimum timing for prognostication. Not surprisingly, in a recent study withdrawal of care was the most important predictor of in-hospital mortality post-TBI, thus forming the basis of self-fulfilling prophecies.⁷⁵ More research is needed in this regard, not only for clinical purposes, but also to enable more objective assessment of the efficacy of neuroprotective therapies under research. Finally, adherence to evidence-based guidelines has been shown to improve outcomes after TBI and decrease hospital expenses.⁷⁶