Hyperosmolar therapies have evolved over the decades, with glycerol, mannitol and various concentrations of saline solutions predominating in clinical care. The mechanism of action of these agents on cerebrohemodynamics is more complex than other agents. Administration of these agents by intravenous infusion leads to an immediate (within seconds to minutes) decrease in blood viscosity. The net effect of this change in blood viscosity is cerebral vasoconstriction associated with unchanged CBF – thereby decreasing cerebral blood volume and causing an immediate decrease in ICP. Secondarily, administration of these agents leads to increases in serum osmolarity. In more delayed time periods (over 20 min to slightly longer), this increase in serum osmolarity leads to net fluid flux from the brain into the serum – leading to decreased edema, relative brain dehydration and decreased ICP.
Similar to other TBI therapies, definitive studies regarding hyperosmolar therapies and outcome are lacking, yet numerous studies point to the beneficial role these agents play in treating children with TBI. James and colleagues reported that a dose of mannitol of 0.5 g/kg resulted in decreased ICP in 78% of patients with a variety of CNS pathologies (e.g., trauma, brain tumor, encephalopathy and subarachnoid hemorrhage), while a larger dose of 1 g/kg was nearly universally effective (99%). Furthermore, Mendelow and colleagues demonstrated similar effects of mannitol on ICP with additional benefit also noted to CPP and CBF. In the previously cited manuscript by Shapiro and Marmarou, mannitol was demonstrated to be effective in improving cerebrohemodynamics by changes in both ICP and PVI. Specifically, PVI increased and ICP decreased in almost all of the 22 children studied (the two nonresponders died from uncontrolled intracranial hypertension shortly after the study). In an ambitious study to attempt to determine if mannitol can improve outcome, Kasoff and colleagues stratified 25 patients based on severity of illness and level of care required (ICP monitoring vs ICP/mannitol vs ICP/mannitol/barbiturates). The authors found that the use of mannitol led to improved outcome as measured by expected outcomes based on trauma scores (88% observed vs 83% predicted).
Complications of mannitol include concerns for hypovolemia due to diuresis, as well as an association between mannitol use and renal failure at extremes of serum osmolarity. Hypertonic saline maintains intravascular volume and has not shown similar renal effects in studies to date. Various concentrations of hypertonic saline have been reported,[31,32] but most studies in children involve administration of 3% saline. Simma and colleagues performed a prospective, randomized, controlled study of hypertonic saline versus normotonic solutions in TBI in children and found: less fluid required for resuscitation; fewer therapies needed for ICP in the first 3 days; decreased duration of mechanical ventilation; and decreased length of hospital stay. More directly, Peterson and colleagues reported successful treatment of intracranial hypertension with 3% saline in a retrospective cohort of 68 children. Of interest, this group demonstrated improved outcomes based on trauma scores and found no adverse events despite extreme hyperosmolarity (serum Na > 180 meq/l in some cases). Khanna and colleagues reported the use of continuous hypertonic saline solution in ten children and found that institution of this therapy resulted in decreased ICP spikes with improved CPP. Finally, in a double-blind, crossover study comparing 3% saline to normal saline for intracranial hypertension, Fisher and colleagues found a 25% decrease in ICP after 3% saline and no change in the normal saline group.
Barbiturates & Drug-induced Coma
Barbiturates and other agents that induce coma have been used as therapy for TBI for decades. Such therapies are administered to decrease metabolic activity of the brain that will lead to decreased blood flow and blood volume, ultimately leading to decreases in ICP. Other effects, such as the change in cerebrovascular tone and inhibition of lipid peroxidation, may also play a role in neuroprotection.[37–39] Inducing coma by a variety of agents, including propofol, benzodiazepines and others, may lead to beneficial effects on cerebrohemodynamics. However, since the large majority of data outlines the use of pentobarbital, this review will focus entirely on this important agent.
Pentobarbital therapy has been associated with several beneficial effects in adult TBI victims, including improved ICP, improved brain oxygenation and decreased markers of excitotoxicity in CSF. However, despite its widespread use, improvements in long-term neurological outcome are lacking for both adults and children. Pittman and colleagues found that pentobarbital therapy led to decreases in ICP in more than 50% of children with TBI who were refractory to other therapies. However, the study design prevented more definitive conclusions regarding pentobarbital's effect on improved outcome. In another relatively large case series, Kasoff and colleagues found that pentobarbital use was closely linked to hemodynamic instability, with nearly all of the 25 children requiring vasoactive medications to maintain blood pressure during pentobarbital therapy. As data supporting barbiturate therapy are relatively scant, the 2003 guidelines for pediatric TBI treatment contain no strategy for its use except to state that 'high-dose barbiturate therapy may be considered in hemodynamically stable patients with salvageable severe head injury and refractory intracranial hypertension'.
Experimental Therapy: Hypothermia
The most promising experimental therapy that may become standard care in the near term is therapeutic hypothermia. Whole-body hypothermia is categorized as moderate when the goal temperature is between 32 and 33°C, and this level of hypothermia has been demonstrated to be neuroprotective for adults with cardiac arrest and neonates with perinatal asphyxia. While the proposed mechanism of action of hypothermia is not entirely understood, it is assumed that hypothermia leads to decreased cerebral metabolism, ultimately leading to decreases in CBF and cerebral blood volume. It is likely that other effects of hypothermia – decreases in enzyme activity, alterations in membrane permeability, preservation of antioxidant reserve, and others – may also have significant neuroprotective effects.
The use of hypothermia for TBI is quite dichotomous at this time. Almost all studies in hypothermia report a decrease in ICP with application of hypothermia, whether it is applied early after TBI as a neuroprotectant or in the delayed phase as a rescue therapy for recalcitrant intracranial hypertension. In a Phase III trial of early hypothermia including 392 adults after TBI, Clifton and colleagues found that there was an 11% difference in ICP over the hypothermia period between normothermic and hypothermic adults. Similarly, Zhi and colleagues found almost a 20% decrease in ICP between hypothermic and normothermic groups in their large Phase III study. In children, Adelson and colleagues report a 19.6% difference in ICP between hypothermic children and those who remained normothermic. Owing to this compelling data, it appears logical to offer hypothermia as a rescue therapy when other therapies have failed.
However, the application of hypothermia as a first-line therapy to improve outcome has been less successful. After being a standard therapy decades ago, interest in hypothermia in trauma was rekindled by Marion and colleagues who demonstrated an improved neurological outcome at 6 months in adults randomized to hypothermia (32–33°C for 24 h). A multi-center trial by Clifton and colleagues sought to confirm this single-center study, but failed to demonstrate any significant beneficial effect. Proponents of hypothermia point out that there were significant treatment variations within the 4-site consortium that may have played a crucial role in this negative trial.
Since the Marion study demonstrated that younger age was associated with an even greater treatment effect for hypothermia, there was enthusiasm to determine if early hypothermia could be an effective neuroprotectant in children. Adelson and colleagues performed a Phase II safety study in 48 children and found that hypothermia was safe and there was a trend toward a decrease in mortality at 3 months. However, in a Phase III trial, Hutchison and colleagues have recently reported a negative study of early hypothermia that was carried out by the Hypothermia Pediatric Head Injury Trial Investigators and the Canadian Critical Care Trials Group. In this study, 225 children were randomized and a detrimental effect of hypothermia was observed (31% poor outcome in the hypothermia group and 22% poor outcome in the normothermia group). Questions about the applicability of the study were raised because the outcome tested (a composite outcome score was used at 3 months by phone interview of parents) was of questionable reliability, there was a substantial amount of hypotension reported in the hypothermia group that may have affected the outcome and there was a profound degree of hyperventilation in both groups of the study that may not reflect current practice in all centers. As a result of this trial, it is possible that hypothermia treatment offers all-risk-and-no-reward for children after TBI and its use should be abandoned. Alternatively, it is possible that the particular protocol tested in this ambitious and arduous trial could be altered to provide safer neuroprotection without incurring substantial risks to the child. This hypothesis is currently being tested in an ongoing clinical trial (Pediatric Traumatic Brain Injury Consortium: Hypothermia – also known as the 'Cool Kids Trial').
Pediatr Health. 2009;3(6):533-541. © 2009 Future Medicine Ltd.
Cite this: Emergency Treatment Options for Pediatric Traumatic Brain Injury - Medscape - Dec 01, 2009.