SECTION EDITOR: DAVID S. WARNER.
In a model of closed head trauma (CHT) in rats, we reported that the IV administration of 250 mL/kg of 5% dextrose in water (D5W) increased the concentration of glucose in blood, increased brain edema in the noncontused hemisphere, and worsened the neurological severity score (NSS) and mortality rate four hours after CHT . In contrast, the IV administration of 250 mL/kg of 0.9% saline (NS) or lactated Ringer's solution caused no significant change in blood glucose, brain edema, NSS, or mortality rate at four hours compared with untreated rats [1,2]. Based on these studies, it was not certain whether the worsening of NSS and mortality rate seen with D5W resulted from brain edema or from increased glucose.
Accordingly, we designed another study to examine the effects of the IV administration of 250 mL/kg of 5% dextrose in 0.9% saline (D5NS) and 0.45% saline (1/2 NS). We hypothesized that D5NS would increase blood glucose but, being hyperosmolar with respect to blood, would not increase brain edema. Conversely, 1/2 NS should not increase blood glucose but, being hypoosmolar with respect to blood, should increase brain edema. We found that D5NS decreased brain edema, increased blood glucose, and caused no significant change in NSS or mortality rate . We also found that 1/2 NS increased the mortality rate and caused no significant change in blood glucose, brain edema, or NSS.
Because both D5W  and D5NS  increase blood glucose, but only D5W  increases edema and worsens outcome, we concluded that worsening of outcome may be caused more by edema formation (regardless of whether glucose is concomitantly increased) than by increased glucose per se. An association between hyperglycemia and poor neurological outcome has been reported in both adult and pediatric head-injured patients and experimental models of ischemia . Regarding the comparison between NS and 1/2 NS in our earlier studies [1-3], because 1/2 NS worsened the mortality rate but did not cause cerebral edema at four hours as expected, these data did not address the question of whether an increase in blood glucose is required to worsen the NSS and mortality rate associated with cerebral edema formation after the IV administration of D5W after CHT.
In a study designed to examine the time of occurrence of the greatest change in the NSS, brain edema, and blood-brain barrier permeability after CHT in rats in the absence of IV fluid administration, we measured brain tissue specific gravity and water content before CHT and 15 minutes; 1, 2, 4, 10, and 24 hours; and 2, 4, and 7 days after CHT . The graph of the data gives the appearance of an initial increase in brain edema at 2-4 hours, followed by a second larger increase in brain edema at 24 hours. Thus, it is possible that although brain edema was not demonstrable at four hours when 1/2 NS was administered IV after CHT, brain edema may be demonstrable at two hours. Accordingly, the present study was designed to examine brain edema two hours after CHT when each of the four above-mentioned IV fluids was administered.
This study was approved by the animal care committee of the Ben-Gurion University of the Negev, Beer-Sheva, Israel. Ninety adult Sprague-Dawley rats, 238 +/- 32 g (mean +/- SD), were used in this study. By random assignment, 45 rats were allocated to the groups undergoing scalp incision (Groups 1-5) and 45 rats were allocated to the groups undergoing scalp incision followed by CHT (Groups 6-10). Experimental conditions for the 10 groups are summarized in Table 1. This model has been previously described [1,2,6,7]. Briefly, 1.5%-2.0% halothane in oxygen was administered in concentrations sufficient to abolish corneal reflexes. A midline scalp incision was made, and the scalp and underlying muscles were reflected laterally. A cranial blow was then delivered (in the appropriate groups) at a prefixed point over the left hemisphere 1-2 mm lateral to the midline on the skull convexity. The blow was delivered by a free-falling plate, from the center of which protruded a silicone-tipped rod that impacted the skull. Settings on the frame controlled the distance of the fall of the platform. The energy imparted to the skull by the stereotaxically guided plate is directly and linearly related to the distance of the fall, and the nonpenetrating blow causes reproducible brain injury and deterioration of neurologic status [1,2,6,7]. After scalp incision with or without CHT, anesthesia was discontinued, and animals were returned to their cages and supplied with unlimited food and water.
One hour after CHT or scalp incision, the NSS was determined. Rats were then reanesthetized with halothane and placed in the supine position. The femoral region was infiltrated with 2% lidocaine, and 23-gauge catheters were inserted into a femoral artery and vein. A 300-[micro sign]L sample of arterial blood was obtained, and continuous measurement of blood pressure was begun. At 75 min after CHT or scalp incision, IV fluid infusion was begun as appropriate. No IV fluids were given to Groups 1 and 6. The other eight groups received 250 mL/kg of IV fluid over 30 min. At 105 min after CHT or scalp incision, IV fluid infusion was discontinued and a blood sample was obtained. Two hours after CHT or scalp incision (i.e., 15 min after completion of IV fluid infusion), a blood sample again was obtained, and all rats were killed.
The NSS was developed to assess the clinical condition of the rats after CHT [1,2,6,7]. Points are assigned for motor function and behavior. The detailed criteria for scoring have been published previously [1,2,6,7]. The NSS directly measures the deterioration of observable neurological status so that a low score represents nearly intact neurological status (minimum 0) and a high score represents severe neurological dysfunction (maximum 25). The NSS was measured 1 h after CHT or scalp incision by a blind observer.
Rats were decapitated 2 h after CHT or scalp incision, and the entire brain was immediately removed and placed on a frozen plate. In the CHT groups, brain tissue samples weighing up to 50 mg were cut from areas adjacent to the zone of maximal macroscopic damage in the left hemisphere and from the corresponding area in the right hemisphere. In groups with no CHT, brain tissue samples were taken from the corresponding left and right hemisphere areas. These small brain tissue samples were used to determine specific gravity. Specific gravity was determined using linear gradient columns of kerosene and bromobenzene . A calibrated curve was generated for each column using anhydrous K2 SO4 solutions of known specific gravity (1.045, 1.040, 1.035, and 1.025).
Systolic and diastolic arterial blood pressures were measured via the 23-gauge catheter in the femoral artery. Mean arterial blood pressure (MAP) was determined by electronic integration of the systolic and diastolic arterial blood pressures. Blood pressure measurement began 1 h after CHT or scalp incision and was discontinued 2 h after CHT or scalp incision. Blood was sampled from the arterial catheter at 1, 1.75, and 2 h to determine plasma osmolality and blood gas tensions and concentrations of urea, glucose, sodium, and potassium.
Blood pressure, brain tissue specific gravity, and laboratory results are expressed as the mean +/- SD. These data were compared using analysis of variance followed by post hoc evaluation using the Student-Newman-Keuls test. The NSS values were tabulated as median (range) and were compared using the Kruskal-Wallis test with post hoc evaluation using the Mann-Whitney U-test. Mortality rates were compared using Fisher's exact test. A probability value of <0.05 was considered significant.
(Table 2) summarizes the specific gravity of cortical slices taken from the injured and the corresponding contralateral hemisphere in the 10 experimental groups. Within-group comparisons indicated that in all groups with CHT, the specific gravity was decreased (water content was increased) in the injured hemisphere compared with the specific gravity in the noninjured (right) hemisphere. Between-group comparisons indicated that specific gravity of the injured hemispheres was decreased compared with that in both hemispheres in rats with no CHT. Among the groups with CHT, specific gravity in the injured hemispheres of rats given 1/2 NS (Group 7) or D5W (Group 9) was significantly decreased compared with the injured hemispheres of rats given D5NS (Group 8) or no IV fluid (Group 6). Specific gravity in the uninjured hemispheres of CHT rats given D5NS (Group 8) was significantly increased compared with the uninjured hemispheres of the other groups of CHT rats. Among the groups with CHT, the specific gravity in both hemispheres of rats given 1/2 NS (Group 2) or D5W (Group 4) was significantly decreased compared with that in rats given D5NS (Group 3), NS (Group 5), or no IV fluid (Group 1).
There was no significant difference in blood pressure among the five groups with no (Groups 1-5). There was also no significant difference in blood pressure among the five groups with CHT (Groups 6-10). In Groups 1-5, the MAP was 91 +/- 11 mm Hg (combined value for all groups). In Groups 6-10, the MAP increased to 104 +/- 11 mm Hg (combined value for all groups).
Values from blood sampled after initial insertion of the femoral artery catheter were: plasma osmolality 280 +/- 7 mOsm/kg; blood concentrations of glucose 130 +/- 5 mg/dL, sodium 135 +/- 6 mEq/L, potassium 3.1 +/- 0.1 mEq/L, and urea 25.8 +/- 2.6 mg/dL; and hematocrit 40% +/- 2%. Plasma osmolality was decreased by 1/2 NS and increased by D5NS. Blood glucose was increased by D5NS and D5W; blood sodium was decreased by 1/2 NS, D5NS, and D5W; and blood potassium was decreased by 1/2 NS and D5NS Table 3.
The NSS 1 h after CHT was 18 (15-21) with no IV fluid administration (Group 6), 18 (14-21) with 1/2 NS (Group 7), 20 (17-22) with D5NS (Group 8), 16 (14-19) with D5W (Group 9), and 17 (15-19) with NS (Group 10). The NSS did not differ significantly among groups.
The principal finding of the present study is that the IV administration of 250 mL/kg of 1/2 NS or D5W after CHT caused significantly greater cerebral edema, as indicated by a significantly greater decrease of brain tissue specific gravity in the injured hemisphere, than did the IV administration of D5NS or NS or no fluid administration. The finding of greater cerebral edema at two hours with 1/2 NS and D5W permits a more complete interpretation of our previous studies [1-3] and addresses the question of whether outcome from CHT is worsened by cerebral edema only when blood glucose is concomitantly increased. In our previous studies, both D5W and D5NS increased blood glucose, but only D5W increased edema and worsened outcome after CHT [1-3]. Although we speculated that worsening of outcome may be caused more by edema formation (regardless of whether glucose is concomitantly increased) than by increased glucose per se, that conclusion could not be confirmed because 1/2 NS worsened the mortality rate but did not cause cerebral edema at four hours as expected. The finding in the present study that 1/2 NS caused cerebral edema at 2 hours confirms that the worsened outcome after CHT (mortality rate 50%) previously reported at 24 hours may relate to postraumatic cerebral edema without a concomitant increase in blood glucose .
The present study also provides information about changes in factors two hours after CHT that may affect edema formation. This two-hour data complements previously reported data on changes in factors four hours after CHT that may affect edema formation [1-3]. We previously reported that at four hours, D5W caused no significant change in blood osmolality and decreased the blood sodium concentration; NS caused no change in osmolality or sodium; D5NS increased osmolality and decreased sodium; and 1/2 NS decreased osmolality and sodium. In the present study, we found that at two hours, all four IV fluids produced the same changes in osmolality and sodium as at four hours. Our findings of worsening cerebral edema two hours after CHT when D5W or 1/2 NS were administered are consistent with previous reports that edema formation is inversely related to the increase or decrease from normal values of blood osmolality and sodium concentration .
Several of the findings of the present study are consistent with previously reported results and support the validity of this model. In the present study, the lower brain tissue specific gravity in rats given 1/2 NS or D5W but not subjected to CHT is consistent with the time course of equilibration of hypoosmolar solutions across the blood-brain barrier [5,9]. The greater brain tissue specific gravity in the uninjured hemisphere of rats with CHT that received D5NS is consistent with the well known effect of hyperosmolar solutions to decrease brain water content [9,10]. The lower brain tissue specific gravity in the injured hemisphere of all groups of rats with CHT is consistent with previous reports of rapidly developing cerebral edema after cranial impact . The NSS at one hour are consistent with values previously reported for this model and indicate that an adequate cranial blow was delivered [1-3,11,12].
Limitations of the present study are that the rats breathed spontaneously during and after CHT, that systemic arterial blood pressure was not measured during the 0- to 60-minute period after CHT, and that brain temperature was not measured. Two consequences of spontaneous ventilation that might affect the present outcome measures are hypoxemia and hypercapnia. Because PaO2, PaCO2, and blood pressure were not measured during the 0- to 60-minute period after CHT, there are no data to prove that hypoxemia, hypercapnia, and hypotension did not occur. However, hypoxemia, hypercapnia, and hypotension severe enough to affect outcome measures might also be expected to have affected the NSS one hour after CHT. In the present study, the NSS at one hour did not differ among groups. PaO2, PaCO2, and MAP values did not differ among groups during the 60- to 120-minute period after CHT. In a later study using this same model, MAP was monitored during CHT and the 0- to 60-minute period after CHT and did not differ between groups . In another subsequent study using this same model, temporalis muscle temperature (and rectal temperature) were measured during CHT and 1, 4, 24, and 48 hours after CHT and did not differ between groups .
In summary, in our previous studies, both D5W and D5NS increased blood glucose, but only D5W increased edema and worsened outcome. These results suggested, but did not prove, that worsening of outcome after CHT may be caused more by edema formation than by increased glucose per se. The results of the present study and those of our previous studies indicate that 1/2 NS given after CHT increases edema and worsens outcome without altering blood glucose. Further, NS or lactated Ringer's solution does not alter edema, blood glucose, or outcome. The present results support the proposal that edema is a greater risk factor for worsening of outcome after CHT than is increased blood glucose.
1. Shapira Y, Artru AA, Qassam N, et al. Brain edema and neurological status with rapid infusion of 0.9% saline or 5% dextrose following head trauma. J Neurosurg Anesthesiol 1995;7:17-25.
2. Feldman Z, Zachari S, Reichenthal E, et al. Brain edema and neurological status with rapid infusion of lactate Ringer's or 5% dextrose solution following head trauma. J Neurosurg 1995;83:1060-6.
3. Gurevich B, Talmor D, Artru AA, et al. Brain edema, hemorrhagic necrosis volume, and neurological status with rapid infusion of 0.45% saline or 5% dextrose in 0.9% saline following closed head trauma in rats. Anesth Analg 1997;84:554-9.
4. Avellino AM, Lam AM, Winn HR. Management of acute head injury. In: Albin MS, ed. Textbook of neuroanesthesia: with neurosurgical and neuroscience perspectives. New York: McGraw-Hill, 1997:1137-75.
5. Shapira Y, Setton D, Artru AA, Shohami E. Blood-brain barrier permeability, cerebral edema, and neurological function after closed head injury in rats. Anesth Analg 1993;77:141-8.
6. Shapira Y, Artru AA, Cotev S, et al. Brain edema and neurological status following head trauma in the rat. Anesthesiology 1992;77:79-85.
7. Shapira Y, Shohami E, Sidi A, et al. Experimental closed head injury in rats: mechanical, pathophysiologic, and neurologic properties. Crit Care Med 1988;16:258-65.
8. Zornow MH, McQuitty C, Prough DS. Perioperative fluid management of the neurosurgical patient. In: Albin MS, ed. Textbook of neuroanesthesia: with neurosurgical and neuroscience perspectives. New York: McGraw-Hill, 1997:117-48.
9. Zornow MH, Scheller MS. Intraoperative fluid management during craniotomy. In: Cottrell JE, Smith DS, eds. Anesthesia and neurosurgery. St. Louis: Mosby-Year Book, 1994:247-59.
10. Todd MM, Tommasino C, Moore S. Cerebral effects of isovolemic hemodilution with a hypertonic saline solution. J Neurosurg 1985;63:944-8.
11. Feldman A, Gurevich B, Artru AA, et al. Magnesium given one hour following head trauma decreases brain edema and improves neurological outcome. J Neurosurg 1996;85:131-7.
12. Shapira Y, Gurevich B, Artru AA, et al. Influence of closed head injury on isoflurane MAC in the rat. J Neurosurg Anesthesiol 1997;9:51-7.
13. Shapira Y, Talmor D, Artru AA, et al. Induced hypertension and neurological outcome after head trauma [abstract]. J Neurosurg Anesthesiol 1997;9:390.
14. Gurevich B, Artru AA, Lam AM, et al. Effects of a new NMDA receptor antagonist, NPS 846, on neurological outcome, following closed head trauma in rats [abstract]. Anesth Analg 1997;84:S357.