Hypertonic saline for cerebral edema and elevated intracranial pressure JOSE´ I. SUAREZ, MD erebral edema and elevated intracranial movements. Because transport through the BBB is a pressure (ICP) are important and frequent selective process, the osmotic gradient that a particle problems in the neurocritically ill patient. can create is also dependent on how restricted its They can both result from various insults permeability through the barrier is. This restriction is C expressed in the osmotic reflection coefficient, which to the brain. Improving cerebral edema and decreas- ing ICP has been associated with improved out- ranges from 0 (for particles that can diffuse freely) to come.1 However, all current treatment modalities 1.0 (for particles that are excluded the most effec- are far from perfect and are associated with serious tively and therefore are osmotically the most active). adverse events:1–4 indiscriminate hyperventilation The reflection coefficient for sodium chloride is can lead to brain ischemia; mannitol can cause 1.0 (mannitol’s is 0.9), and under normal conditions intravascular volume depletion, renal insufficiency, sodium (Na+) has to be transported actively into the and rebound ICP elevation; barbiturates are associat- CSF.5,6 Animal studies have shown that in condi- ed with cardiovascular and respiratory depression tions of an intact BBB, CSF Na+ concentrations and prolonged coma; and cerebrospinal fluid (CSF) increase when an osmotic gradient exists but lag drainage via intraventricular catheter insertion may behind plasma concentrations for 1 to 4 hours.5 result in intracranial bleeding and infection. Thus, elevations in serum Na+ will create an effec- Other treatment modalities have been explored, tive osmotic gradient and draw water from brain into and hypertonic saline (HS) solutions particularly the intravascular space. appear to be an appealing addition to the current therapeutic avenues for cerebral edema. This article Cerebral edema and intracranial dynamics succinctly reviews some of the basic concepts and Cerebral edema is defined as an increase in brain 7 mechanisms of action of HS and discusses some of water leading to an increase in total brain mass. its possible clinical applications. There are three major categories of brain edema: • Vasogenic edema, which is caused by increased ■ PHYSIOLOGIC CONTEXT permeability of the endothelial cells of brain capil- laries and is seen in patients with brain neoplasms The blood-brain barrier • Cytotoxic edema, which results from the influx of The blood-brain barrier (BBB) represents both an water into cells. This type of edema may be caused anatomic and a physiologic structure. The BBB is by energy depletion with failure of the ATP-depen- made up of tight junctions between the endothelial dent Na+-K+ pump (ie, cerebral infarction) or low cells of the cerebral capillaries.5 Various mechanisms extracellular Na+ content (ie, hyponatremia). exist for compounds to cross the BBB, including • Interstitial edema, in which CSF diffuses through active transport, diffusion, and carrier-mediated the ependymal lining of the ventricles into the periventricular white matter. This type of edema is seen with hydrocephalus. From the Department of Neurology and Neurosurgery, Uni- versity Hospitals of Cleveland and Case Western Reserve It is important to point out that different types of University, Cleveland, Ohio. edema can coexist in the same patient. For instance, Address: Jose´ I. Suarez, MD, Department of Neurology, Uni- brain ischemia is associated with both cytotoxic and versity Hospitals of Cleveland, 11100 Euclid Avenue, vasogenic edema. Cleveland, OH 44106; e-mail:[email protected]. The presence of cerebral edema, with the subse- CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 71 • SUPPLEMENT 1 JANUARY 2004 S9 SUAREZ ■ HYPERTONIC SALINE quent increase in brain mass, alters the intracranial • Restoration of normal membrane potentials contents (brain, blood, and CSF). Small increases through normalization of intracellular sodium and in brain volume can be compensated by changes in chloride concentrations.13 CSF volume and venous blood volume. Beyond • Reduction of extravascular lung volume, lead- 15 that, changes in intracranial volume (ΔICV) will ing to improved gas exchange and improved PaO2. result in changes in ICP (ΔICP), which has been termed compliance (ΔICV/ΔICP). When brain ■ EXPERIMENTAL SUPPORT FOR THE EFFICACY compliance decreases, such as when intracranial OF HYPERTONIC SALINE volume rises, ICP rises.8 However, it is important to realize that focal cerebral edema can create ICP gra- Animal studies dients and cause tissue shifts in the absence of a HS solutions have been studied extensively in a global increase in ICP.1 variety of animal models, as thoroughly detailed in a recent review.13 The literature suggests that fluid ■ HYPERTONIC SALINE: MECHANISMS OF ACTION resuscitation with an HS bolus after hemorrhagic shock prevents the ICP increase that follows resus- HS solutions can possibly affect the volume of the citation with standard colloid and crystalloid fluids intracranial structures through various mechanisms. for 2 hours or less. This effect can be maintained for All or several of them are likely to be interacting to longer periods by using a continuous HS infusion. achieve the end result of HS therapy: reduction of HS may be superior to colloid solutions with regard cerebral edema and elevated ICP. These mecha- to ICP response during the initial period of resusci- nisms are summarized below: tation.16,17 In animal models of cerebral injury, the • Dehydration of brain tissue by creation of an maximal ICP-reducing effect of HS is appreciated osmotic gradient, thus drawing water from the with focal lesions, such as cryogenic injury or intra- parenchyma into the intravascular space. As men- cerebral hemorrhage. Again, the ICP reduction may tioned above, this would require an intact BBB. be caused by reduction in water content in areas of Experimental evidence suggests that the brain the brain with the BBB intact, such as the nonle- water–reducing properties of HS are accomplished sioned hemisphere and the cerebellum. HS has also at the expense of the normal hemisphere. been compared with mannitol and was found to • Reduced viscosity. HS solutions enhance intra- have equal efficacy in reducing ICP but to have a vascular volume and reduce viscosity.9 The autoreg- longer duration of action and to yield greater ulatory mechanisms of the brain vasculature have improvement in cerebral perfusion pressure.13 been shown to respond not only to changes in blood pressure but also to changes in viscosity.10 Thus, a Human studies decrease in blood viscosity results in vasoconstric- Despite the numerous studies in animal models, tion in order to maintain a stable cerebral blood most of the evidence in humans is based on the pub- flow (CBF). lication of case series and a few randomized studies. • Increased plasma tonicity. It has been postulated, Some of the published studies are briefly reviewed based on experimental animal data, that increased here. Readers are referred to a recent review13 for a plasma tonicity, such as that seen after HS adminis- more detailed description. tration, favors more rapid absorption of CSF.11 Acute ischemic stroke. HS in two different con- • Increased regional brain tissue perfusion, possi- centrations, 7.5% and 10%, has been used to reduce bly secondary to dehydration of cerebral endothelial ICP in patients after large cerebral infarcts.18,19 cells and erythrocytes, facilitating flow through cap- Schwarz et al18 compared the effect of 100 mL of illaries.12 7.5% HS hydroxyethyl starch (osmolarity 2,570 • Increased cardiac output and mean arterial blood mosm/L) and 200 mL of 20% mannitol (osmolarity pressure, with resultant augmentation of cerebral 1,100 mosm/L) in 9 patients with stroke randomized perfusion pressure, most likely due to improvement to one of the two treatments. HS hydroxyethyl of plasma volume and a positive inotropic effect.9,13 starch caused a greater and earlier peak reduction in • Diminished inflammatory response to brain ICP, although mannitol caused more improvement injury, which has been demonstrated with HS in cerebral perfusion pressure. These same administration.14 researchers studied the effect of 10% saline bolus S10 CLEVELAND CLINIC JOURNAL OF MEDICINE VOLUME 71 • SUPPLEMENT 1 JANUARY 2004 SUAREZ ■ HYPERTONIC SALINE infusions in 8 patients in whom mannitol had have reported on the use of HS in patients with var- failed.19 HS reduced ICP by at least 10% in all the ious intracranial pathologies.34–37 instances it was used, and the maximal effect was Gemma et al34 performed a prospective, random- noted at 20 minutes after the end of the infusion. ized comparison of 2.5 mL/kg of 20% mannitol and Even though ICP rose subsequently, it did not reach 7.5% saline in patients undergoing elective supra- pretreatment values during the 4 hours of data tentorial procedures. They found that the two treat- recording. ments had similar effects on CSF pressure and on Intracranial hemorrhage. There has been one clinical assessment of brain bulk. However, the report of 2 patients with nontraumatic (presumably administered solutions used were not equiosmolar. hypertensive) intracranial hemorrhage who were In a retrospective study, Qureshi et al35 determined treated with continuous HS infusion.20 Both the effect of continuous 3% saline/acetate infusion patients improved clinically after 24 hours of treat- on ICP and lateral displacement of the brain in ment but deteriorated at 48 and 96 hours despite patients with cerebral edema and a variety of under- continued HS infusion. Repeat CT scanning lying cerebral lesions. The authors found a reduction showed extension of edema. These findings were in mean ICP within the first 12 hours, correlating attributed to a rebound effect similar to that with an increase in serum sodium concentration, in described with mannitol. patients with traumatic brain injury and postopera- Subarachnoid hemorrhage. Two studies have tive edema, but not in patients with nontraumatic been published on the effect of HS on clinical intracranial hemorrhage or cerebral infarction.