Effects of epidural hypothermic saline infusion on locomotor outcome and tissue preservation after moderate thoracic spinal cord contusion in rats

Transcription

1 J Neurosurg Spine 2: , 2005 Effects of epidural hypothermic saline infusion on locomotor outcome and tissue preservation after moderate thoracic spinal cord contusion in rats CARLOS E. CASAS, M.D., LOREN P. HERRERA, M.D., CHAD PRUSMACK, M.D., GLADYS RUENES, ALEXANDER MARCILLO, M.D., AND JAMES D. GUEST, M.D., PH.D., F.R.C.S.(C) The Miami Project to Cure Paralysis and the Department of Neurological Surgery, University of Miami, Florida Object. Regionally delivered hypothermia has advantages over systemic hypothermia for clinical application following spinal cord injury (SCI). The effects of local hypothermia on tissue sparing, neuronal preservation, and locomotor outcome were studied in a moderate thoracic spinal cord contusion model. Methods. Rats were randomized to four treatment groups and data were collected and analyzed in a blinded fashion. Chilled saline was perfused into the epidural space 30 minutes postcontusion to achieve the following epidural temperatures: C (16 rats), C (13 rats), and C (13 rats). Hypothermia was continued for 3 hours when a 45-minute period of rewarming was instituted. In a fourth group a moderate contusion only was induced in 14 animals. Rectal (core) and T9 10 (epidural) temperatures were measured continuously. Locomotor testing, using the Basso-Beattie- Bresnahan (Ba-Be-Br) scale, was performed for 6 weeks, and rats were videotaped for subsequent analysis. The lesion/preserved tissue ratio was calculated throughout the entire lesion cavity and the total lesion, spinal cord, and spared tissue volumes were determined. The rostral and caudal extent of gray matter loss was also measured. At 6 weeks locomotor recovery was similar in all groups (mean Ba-Be-Br Scale scores , , , and [p = 0.77] for all four groups, respectively). No significant differences in spared tissue volumes were found when control and treatment groups were compared, but gray matter preservation was reduced in the infusion-treated groups. Conclusions. Regional cooling applied 30 minutes after a moderate contusive SCI was not beneficial in terms of tissue sparing, neuronal preservation, or locomotor outcome. This method of cooling may reduce blood flow in the injured spinal cord and exacerbate secondary injury. KEY WORDS spinal cord injury hypothermia neuroprotection ischemia rat T HE role of hypothermia as a neuroprotective strategy to reduce the extent of SCI has been controversial since the first promising results were reported in the 1950s, 50, s, 18,72 and 1970s. 23,27 Two principal methods have been used to cool the spinal cord total body cooling (systemic hypothermia) and local spinal cord cooling. 29 The practice of systemic hypothermia derives from studies in which profound cooling was shown to extend the duration of ischemic tolerance of the brain and other organs, including the spinal cord; 54 however, deep systemic hypothermia is associated with potentially life-threatening complications such as ventricular arrhythmia 52,59 and impaired blood coagulation. 67 Local cooling came to prominence because of the studies of Albin, et al., 1,2 who conducted experiments in spinal cord injured monkeys in which they used profound levels of local cooling (11 C). In the past two Abbreviations used in this paper: ANOVA = analysis of variance; Ba-Be-Br = Basso-Beattie-Bresnahan; FBFVA = frame-byframe video analysis; NYU = New York University; SCBF = spinal cord blood flow; SCI = spinal cord injury; SD = standard deviation; STV = spared tissue volume. 308 decades new evidence has accrued that mild degrees of systemic hypothermia (for example, 32.5 C) can be neuroprotective after brain insults 7,34,42,43 and SCI. 80 Still, there are significant concerns about the safety of subjecting patients with acute SCI even to mild systemic hypothermia. 61 These include increased risks of infection 25,37,78 and possible myocardial ischemia during rewarming. 21 Patients who have sustained acute SCI have altered autonomic function, 39,69 and impaired myocardial function has been demonstrated in experimental studies. 26,68 Patients receiving large steroid doses are relatively immunosuppressed and are at high risk for infections such as pneumonia. 6,49 In addition, because of shivering, it may be difficult to maintain moderate systemic hypothermia in conscious patients with SCI. The ability to maintain a normal core temperature and avoid the risks of systemic hypothermia makes local cooling strategies attractive for clinical application following SCI. Several methods of local cooling have been shown to be neuroprotective in spinal cord ischemia models in several species. 3,13,40,44,53,63,74 Furthermore, even mild degrees of local hypothermia (30 34 C) have been shown to be protective against transient spinal cord ischemia. 45,46,73 A neu-

2 Epidural hypothermic neuroprotection after SCI in rats roprotective effect of local cooling has also been shown in models of spinal cord compression in rats 70 and in canines 35,36 as well as in some SCI studies; 65,66 however, in a study by Dimar and colleagues 17 involving regional cooling to achieve a 19 C epidural temperature following a severe contusive injury, the authors found no benefit in terms of histopathological, behavioral, and electrophysiological outcomes. When an ischemic compressive injury was added to the initial contusion, local cooling was beneficial. Thus, several groups have shown that local cooling provides a neuroprotective effect after diverse spinal compression/ischemic injuries in several species. In clinical practice perfusion cooling of the epidural space has been shown to reduce the risk of paraplegia following aortic aneurysm surgery. 8,9,15 Thus, local cooling is a practical, achievable, and safe strategy, although its efficacy for contusive SCI is unclear. One theoretical advantage of epidural perfusion cooling over other local cooling methods is that it avoids temperature gradients within the spinal cord because the coolant completely surrounds the spinal cord. Clinical application could be as simple as placement of a multilumen central venous catheter into the epidural space, which would avoid the need for cumbersome and complex devices after laminectomy. 48 Because epidural perfusion cooling is effective for spinal cord neuroprotection under conditions of spinal cord ischemia 3,47 and compression and because it has direct potential for clinical translation, we tested its efficacy following a moderate contusive injury (12.5 g/cm) delivered with the NYU injury device. We selected this magnitude of injury because there is considerable preserved tissue after the initial contusion and secondary injury processes substantially propagate the extent of injury. Because the optimal levels of spinal cord cooling are not known, we assessed the effect of three different epidural temperatures on histopathological outcome and Ba- Be-Br locomotor recovery after SCI. We studied the volume of preserved tissue and the extent of gray matter preservation 6 weeks postinjury. In previous experimental studies investigators have used dissimilar methodologies and outcome measures, making comparisons difficult. The development of highly standardized injury devices (for example, the NYU impactor) 79 and outcome assessment scales (Ba-Be-Br locomotor assessment scale) 4 in the field of experimental SCI have enhanced the ability to achieve meaningful comparison of data across studies; 5 however, it remains important to refine the utility of these techniques and explore their limitations. This is particularly true when assessing moderately severe thoracic contusion injuries. In the range to which such contused animals recover (Ba-Be- Br Scale Scores 10 14), the extent of forelimb hindlimb coordination is used quantitatively. We have found that there is considerable interobserver subjectivity when making this particular assessment. Therefore, we developed a video-based Ba-Be-Br assessment technique to enhance quantification of forelimb hindlimb coordination and other details such as the degree of paw rotation. Materials and Methods All rodent experiments were conducted after approval of Animal Care and Use Committee protocols in accordance with the National Institutes of Health policy on the humane care and use of laboratory animals. Prior to starting the experiment, we confirmed that our NYU impactor produced lesions similar in magnitude to those reported by other investigators. 5,79 We performed 12.5-, 25-, and 50- g/cm injuries in three groups of five animals, killed the animals after 24 hours, and sent the frozen spinal cord tissue to the laboratory of Dr. Wise Young for determination of the mean lesion volume. 14 Surgical Procedures For the epidural hypothermia experiment 56 Sprague Dawley female rats (mean weight SD g) were randomly assigned to one of four groups. After induction of anesthesia with halothane for 10 to 15 seconds in a sealed jar, the rats were placed on the surgical table and anesthesia was maintained with 2% halothane, 60% oxygen, and 40% nitrous oxide delivered through a humidified anesthetic circuit. The animal s back was shaved and prepared with betadine solution. An incision was performed to expose the T9 11 laminae. Under a surgical microscope, a complete T-10 laminectomy was performed. The dura mater was preserved to avoid direct saline infusion into the subarachnoid space. The 14 control animals underwent a T-10 laminectomy, and a 12.5-g/cm contusion was induced. The treated animals (Groups I III) received similar injuries, but 30 minutes postinjury their epidural space was perfused with ice-chilled saline solution for 3 hours consecutively. Rewarming was achieved during a 45-minute period. The target epidural temperatures were 20, 30, and 37.5 C. Three additional animals underwent invasive monitoring of intramedullary temperature. Prior to the randomized experiment, six other animals were used to study the relationship between the perfusate flow rate and the epidural temperature. These animals underwent carotid artery cannulation and measurement of blood pressure as well as PO 2, PCO 2, and ph at various epidural temperatures. Injury Model The NYU impactor 24 was used to induce a moderate contusion injury (12.5 g/cm) in the exposed spinal cord. This device allowed us to record the compression distance, compression velocity, and duration of each contusion. These data were used to determine which animals should be rejected from analysis and to establish that intergroup injury magnitudes were similar. Animals in which there was a compression distance greater than two SDs from the mean for their group were excluded from analysis. Cooling Technique Thirty minutes postinjury, 0.9% ice-chilled saline solution was infused into the epidural space by using an infusion pump (model ; Harvard Apparatus, South Natick, MA). The cooling setup is shown in Fig. 1. The temperature of a 50-ml syringe was maintained near 4 C by application of ice. The distal end of an Intramedic polyethylene PE-10 catheter (Becton Dickinson, Parsippany, NJ) was located at the base of the T-11 spinal process. Chilled fluid flowed from T-11 to T-9 where an identical catheter was placed and connected to a suction device. The suction was adjusted so that the dura was constantly immersed in fluid. The epidural temperatures were controlled by varying the infusion rates to (Group I), (Group II), and ml/hour (Group III). The rewarming period was achieved by gradually reducing the infusion rate to zero during a 45-minute period. This setup models an epidural perfusion technique that could be readily applied in patients, 9 conceivably without a need for laminectomy. Subarachnoid perfusion was not used because subdural placement of catheters can cause additional mechanical trauma to the spinal cord (unpublished data). Physiological Data Recording The animal was kept on a homeothermic blanket during the experimental procedure. The blanket s control unit (model NP R; Harvard Apparatus) was connected to a rectal temperature probe and set to maintain the core temperature at 37 C. The distal end of a minihypodermic thermocouple 0.2-mm-diameter probe (model HYP-0; Omega, Stamford, CT) was located between the base of the T-9 spinal process and the posterior surface of the dura to record the local epidural temperature at the injury site (Fig. 1B). Epidural and core temperatures as well as O 2 saturation and heart 309

3 C. E. Casas, et al. I. II. SUMMARY OF CONTUSION PARAMETERS* Treatment Group Variable Control Rats I II III p Value compression (mm) height (mm) velocity (m/second) time (msec) * Values are presented as means SDs. III. COMPARISON OF RECTAL AND EPIDURAL TEMPERATURES* Treatment Group ( C) Temperature I II III p Value rectal epidural * Values are presented as means SDs. FIG. 1. Epidural cooling technique: spinal cord (A); epidural temperature probe (B) immersed in perfusate (C) over the site of contusion (D). The core temperature is monitored with a rectal probe (H), and flow from the syringe pump (G) and suction (F) are regulated to maintain the target epidural temperature (E). II. Contusion parameters. III. Mean rectal and epidural temperatures in the experimental groups. rate were monitored continuously and recorded every 30 minutes. In three animals a second probe was placed into the T-10 parenchyma to a depth of 1.5 mm to study the relative changes in epidural and intraspinal temperatures while infusion rates were varied. Because intraspinal temperature recording is invasive and causes additional SCI we used only epidural temperature recording in the animals studied for locomotor and histopathological outcomes. Behavioral Assessment Open-field locomotor function was evaluated once a week for 6 weeks by two trained observers who used the Ba-Be-Br locomotor rating scale, 4 and the mean of their scores was recorded. The animals were pretrained in the open-field environment before surgery was performed. The evaluators were blinded as to which treatment group each coded rat had been assigned. Digital video was recorded during the Ba-Be-Br sessions and an FBFVA performed in 28 of the animals, representative of all the groups in the study. A tripodmounted Sony DCR-PC100 digital camera recorded video at 30 frames per second. For FBFVA, 30-second videoclips were saved as uncompressed files. Sequential frames were displayed at pixels by using video editing software. The FBFVA allowed us to determine the sequence of limb movements. Each limb was assigned a letter (A, B, C, or D), and a score sheet was used to record each successive limb movement. Every four steps by the animal were defined as a complete stepping cycle, and this cycle was evaluated for coordination. The fulfillment of two criteria were 310

4 Epidural hypothermic neuroprotection after SCI in rats TABLE 1 Summary of the modified Ba-Be-Br scale and the FBFVA Ba-Be-Br Scale* Score Ba-Be-Br Scale FBFVA Ba-Be-Br Scale 10 occasional weight-supported plantar steps, no FL HL coordina- occasional weight-supported plantar steps; 0% FL HL tion coordination 11 occasional weight-supported plantar steps; occasional FL HL occasional weight-supported plantar steps, 0 25% coordination FL HL coordination 12 frequent-to-consistent weight-supported plantar steps & occa- frequent-to-consistent weight-supported plantar steps sional FL HL coordination & 25 50% FL HL coordination 13 frequent-to-consistent weight-supported plantar steps & frequent frequent-to-consistent weight-supported plantar steps FL HL coordination & 50 75% FL HL coordination 14 consistent weight-supported plantar steps; consistent FL HL co- consistent weight-supported plantar steps, % ordination; predominant paw position rotated when it makes FL HL coordination; same paw position criteria or initial contact w/ the surface & just before it is lifted off at consistent FL HL coordination & occasional dorsal end of stance or consistent FL HL coordination & occasional stepping dorsal stepping 15 consistent plantar stepping & consistent FL HL coordination; & consistent weight-supported plantar steps; consistent no toe clearance or occasional toe clearance during forward (100%) FL HL coordination, predominant paw limb advancement; predominant paw position is parallel to the position is parallel to body at initial contact & body at initial contact rotated at liftoff 16 consistent plantar stepping & consistent FL HL coordination du- consistent weight-supported plantar steps; consistent ring gait; & toe clearance occurs frequently during forward FL HL coordination, predominant paw position is limb advancement; predominant paw position is parallel at parallel at initial contact & liftoff initial contact & rotated at liftoff 17 consistent plantar stepping & consistent FL HL coordination du- consistent weight-supported plantar steps; consistent ring gait; & toe clearance occurs frequently during forward FL HL coordination, predominant paw position is limb advancement; predominant paw position is parallel at parallel at initial contact & lift off; trunk instability initial contact & liftoff is present & tail is consistently up 18 consistent plantar stepping & consistent FL HL coordination du- consistent weight-supported plantar steps; coordinatring gait; & toe clearance occurs consistently during forward ed gait, predominant paw position is parallel limb advancement, predominant paw position is parallel at throughout stance, consistent trunk stability, tail initial contact & rotated at liftoff consistently up * FL = forelimb; HL = hindlimb. required to score a stepping cycle as coordinated: 1) for every forelimb step a hindlimb step is taken; and 2) hindlimbs alternate during the cycle. We decided that a minimum of four complete sequential cycles was required to assess for frequency of coordination. The Ba-Be-Br scores in the 10 to 15 range were based on the percentage of coordinated cycles (0% = 10, 0 25% = 11, 25 49% = 12, 50 74% = 13, 75 99% = 14, 100% = 15) observed using FBFVA. Exploratory forelimb behavior was excluded from the analysis. To adjust for the absence of audible toe clearance with this video method, we modified the Ba-Be-Br Scale as shown in Table 1. Histopathological Assessment Forty-five days after SCI, rats were deeply anesthetized using a rat cocktail (ketamine 43 mg/ml, xylazine 8.6 mg/ml, and acepromazine 1.4 mg/ml) at 1.5 ml/kg and perfused transcardially with heparinized isotonic saline, and then by 4% paraformaldehyde (0.2 M phosphate buffered, ph 7.4). Each spinal cord was assigned a code that was kept by one of the authors (G.R). None of the other authors knew which experimental group the coded spinal cord tissue belonged to until the analysis was completed. Following perfusion, the spinal cords were removed and postfixed for 48 hours, after which they were placed in PBS (0.1 M) for 2 additional days. The contusion site was localized under a dissecting microscope and two 15-mm-long blocks of tissue were selected rostral and caudal to this point. The blocks were embedded in paraffin and axially sectioned every 10 m by using a rotating microtome. Sequential sections were placed onto gelatin-coated slides. Sections were stained with H & E and Luxol fast blue. A Neurolucida image analysis system (version 5.04; MicroBrightField, Inc., Williston, VT) was used to trace the spinal cord perimeter and the cavity contours at 200- m successive intervals throughout the 30 mm of tissue to reconstruct spinal cord and cavity volumes (Fig. 2). Spared Tissue Volume Calculation. To avoid the confounding effects of the variable distortion of the injured spinal cord between animals, we decided to use cord/cavity ratios. Three-dimensional serial reconstructions and volumetric calculations were computed in Neurolucida/Neuroexplorer. This program permits summation of spinal cord contour and cavity areas to calculate the total spinal cord (Vol total ) and lesion volumes (Vol lesion ) within the injury region. The percentage of STV was calculated using the following formula, STV (%) = 100 [ Vol lesion /Vol total 100]. The epicenter was considered to be the section with the smallest cord/cavity ratio. Based on the localization of the epicenter the rostral and caudal volumes were also derived and compared. Neuronal Loss. The rostral and caudal extent of gray matter loss was estimated by analyzing the H & E/Luxol fast blue stained sections for the presence of neurons in sequential sections at magnification 400. Scanning from the epicenter rostrally and caudally, the first section with at least three definite preserved neurons was selected to delimit the extent of neuronal loss. A preserved neuron was defined as one with a definite nucleolus and visible perinuclear Nissl substance. Statistical Analysis To test the hypothesis that hypothermic saline perfusion resulted in improved locomotor recovery among the four groups, the Ba-Be- Br Scale scores during the 45-day survival period were compared using one-way ANOVA with Tukey post hoc tests. To test the hypothesis that FBFVA improved the accuracy of the Ba-Be-Br assessment tool we hypothesized that FBFVA would result in less variation in the Ba-Be-Br Scale scores. Weekly scores per treatment group were averaged and SDs calculated for both real-time (conventional) Ba-Be-Br and FBFVA Ba-Be-Br. The F statistic was calculated to test for equivalence of the SDs. To test the hypothesis that hypothermic perfusion resulted in tissue sparing, the area of white matter spared in the epicenter, and the mean values of the total, rostral, and caudal spared tissue volumes were also compared using ANOVA. The rostral and caudal extent of loss of gray matter was compared among the four groups by using ANOVA. Linear regression analyses were used to test for a correlation between the spared 311

5 C. E. Casas, et al. FIG. 3. Graph showing independent validation of the NYU impactor prior to initiation of the experiment. Twenty-four-hour postinjury lesion calculated volumes are shown. FIG. 2. Neurolucida method to map contusion volumes. Upper: The spinal cord and cavity perimeters were drawn at 200 m intervals 45 days after injury. Lower: A three-dimensional solid reconstruction with the lesion cavity in blue. Bar = 1 mm. tissue area at the epicenter and the spared tissue volume across the entire region of a detectable lesion cavity. Statistical significance was set at a probability value less than Results are presented as the means SDs. Results Eighty animals were used in the study, of which the first 15 were used to verify the reliability of our NYU impactor device. Six animals were used in the pilot study to determine the infusion rates needed to attain target temperatures and to assess the impact of local cooling on physiological variables. Fifty-six animals underwent randomization and contusive SCI. Of these, two animals in Group III were rejected because their contusion parameters were greater than two SDs from the mean (abnormally high compression distances of 2.29 and 2.18 mm, respectively). Seven animals died prematurely and were not included in the analysis. Validation of the NYU Device Our device produced consistent injuries similar to those published by other investigators (Fig. 3). Contusion Parameters We compared the intra- and intergroup contusion-related variables to exclude significant differences as confounding variables. As shown in Fig. 1 (section II), there were no significant differences in contusion parameters between the four groups. Epidural and Rectal Temperatures Rectal temperature recording was used to monitor the core temperature. The overall mean rectal temperature during the 3-hour cooling period was C. No significant intergroup differences in rectal temperatures were found among injured and control rats. The mean epidural temperatures achieved during the cooling phase were C in Group I (16 rats), C in Group II (13 rats), and C in Group III (13 rats) (Fig. 1 [III]). We were unable to obtain consistent epidural temperatures below 22 C using this technique. There was a nonlinear relationship between the rate of chilled fluid infusion and epidural temperature (Fig. 4 upper). Blood pressure, PO 2, PCO 2, and ph were not affected by epidural perfusion cooling in the six animals in which it was assessed. Epidural and rectal temperatures were similar in rats that received contusion but not epidural fluid perfusion. Epidural/Intraspinal Temperature Gradients. As the infusion rate was increased, the difference between the epidural and intraspinal temperatures increased with the spinal cord becoming colder than the epidural space (Fig. 4 lower). No significant hydrostatic pressures were created by the perfusion. Behavioral Testing The results of hindlimb locomotor assessment using the standard Ba-Be-Br protocol are shown in Fig. 5 upper. 312

6 Epidural hypothermic neuroprotection after SCI in rats FIG. 4. Graphs. Upper: Nonlinear relationship between rate of chilled epidural saline perfusion and epidural temperatures. Lower: The difference ( ) between epidural and intramedullary temperatures increases with increasing rates of chilled saline perfusion. The scores were similar in all groups ( [control], [Group I], [Group II], [Group III]) (p = 0.77). The FBFVA Ba-Be-Br scores were lower than the real-time scores because 100% forelimb hindlimb coordination was rarely observed using the more detailed method of analysis (Fig. 5 lower). In comparing SDs of real-time Ba-Be-Br and FBFVA Ba-Be-Br scores, we found that in 16 of 22 events the SDs of the real-time scores exceeded those of FBFVA Ba-Be-Br scores. Testing for equivalence of the SDs using the F statistic demonstrated that six of 16 instances had significant differences (p 0.05). Conversely, Ba-Be-Br real-time SDs were less than those of FBFVA SDs in five of 22 events with no statistically significant F value. In two instances the SDs were equivalent. This means that FBFVA reduces variation in the assigned Ba-Be-Br scores, presumably because of enhanced accuracy. Histopathological Assessment FIG. 5. Graphs. Upper: Locomotor behavioral outcomes. The Ba-Be-Br score (BBB) was evaluated each week for 6 weeks following the contusion injury. Values are presented as the means SDs. Lower: Comparison of the overall mean value and SD of conventional Ba-Be-Br and FBFVA Ba-Be-Br scores for each experimental group. Figure 6 shows representative epicenter sections from the experimental groups. In some of the sections, necrotic tissue strands were evident within the cavity and this tissue was counted as part of the lesion. A polymorphonuclear inflammatory cell reaction and macrophages were found adjacent to the cavity in almost all of the epicenter sections, but their quantification was not an end point of this study. Qualitatively, no obvious differences were seen between controls and perfusion-cooled animals. An example of a section with three intact neurons is shown in Fig. 7 to illustrate how we made this determination. The preserved tissue area at the epicenter and the total, rostral, and caudal STVs showed nonsignificant differences when comparing the control and treated groups 45 days postinjury (Fig. 8A D). The longitudinal extent of gray matter loss (Fig. 8E) was m (in control), m (in Group I), m (in Group II), and m (in Group III rats). The rostral caudal extent of neuronal loss indicated that gray matter preservation was reduced in the perfusion-cooled animal groups 313

7 C. E. Casas, et al. FIG. 6. Representative histological sections from spinal cord epicenters. Control (A and B), Group I (C and D), Group II (E and F), and Group III (G and H) H & E and Luxol Fast blue. Bar = 0.5 mm. compared with controls (p = 0.05). The mean rostral caudal loss of gray matter following a 12.5-g/cm lesion at 45 days in this study was m. Therefore, a 30- mm block is adequate to characterize the entire cavity if centered on the epicenter. In all cases the first neurons to be observed within preserved gray matter, rostral or caudal to the lesion were motor neurons. Sections further rostral or distal contained larger numbers of preserved neurons. The relative preservation of gray matter in the control group did not correlate with an improved Ba-Be- Br score. Linear regression analysis showed a high correlation between the spared tissue area at the epicenter and the STV over the entire extent of the lesion cavity (r = 0.83, p ). Discussion Failure of Epidural Perfusion Cooling to Provide Neuroprotection Pathomechanisms sensitive to temperature alterations 314

8 Epidural hypothermic neuroprotection after SCI in rats between the two studies are almost identical. Our findings also differ from those of Albin, et al., 1 who used cooler temperatures in primates than we could achieve with epidural perfusion in rats. Our findings are consistent with those of Dimar, et al., 17 who, however, used a more severe injury magnitude and a less rigorous histological analysis. To understand these discrepant results, the following issues should be considered. 1) Local spinal cord perfusion cooling may cause reduced SCBF exacerbating ischemia, whereas systemic hypothermia may not have this effect. 2) Local and systemic hypothermia may affect different pathomechanisms. One effect that may differ between local and systemic cooling is the impact on inflammation, specifically entry of polymorphonuclear leukocytes into the injured spinal cord. Polymorphonuclear leukocytes propagate secondary injury, and their ingress to the injured spinal cord has been reported to be reduced when systemic hypothermia is delivered after SCI. 10 Because the polymorphonuclear leukocytes are constantly rewarmed as they circulate under conditions of local hypothermia, they may still contribute to secondary injury. Value of FBFVA of Open-Field Locomotion Our observed Ba-Be-Br scores after the 12.5-g/cm injury are generally similar to those reported by other investigators; 79 however, analysis of FBFVA Ba-Be-Br score results consistently showed reduction in the assigned Ba-Be-Br score. We think this is due to more accurate assessment of forelimb hindlimb coordination and also of events such as paw rotation that we find difficult to assess in real time. This analysis is particularly valuable in assessing the g/cm injury where recovery to varying extents of forelimb hindlimb coordination is frequently observed. FIG. 7. Method used to assess gray matter preservation. Upper: The first section (from a series of 10- m sequential sections) with three visible intact neurons (white box). Neurolucida reconstruction. Lower: View of the three neurons in the white box (asterisks). H & E and Luxol Fast blue, original magnification 10 (upper); 400 (lower). following ischemic and traumatic injury include glutamate release, stabilization of the blood brain barrier, O 2 radical production, nitric oxide metabolism, protein synthesis, energy metabolism, integrity of membrane function, inflammation, activation of intracellular protein kinases, cytoskeletal breakdown, apoptotic cell death, and gene expression. 11,12,16,30,38,41,60,75,77,82 The neuroprotective effects of hypothermia are linked to the magnitude and duration of the insult, the level of hypothermic temperature, the timing of its application, 32,46,51,64,71 and rewarming. Despite evidence that hypothermia can reduce the severity of several pathomechanisms after central nervous system trauma or ischemia, we found no beneficial effect. This result contrasts with a beneficial result reported by investigators who used a similar injury magnitude but conducted systemic hypothermia. 80 The injury parameters Increase in Epidural/Intraspinal Temperature Gradient With Increased Rates of Epidural Perfusion Because cooling reduces the metabolic demand of neural tissue, it seems reasonable that cooling the spinal cord might improve the survival of injured spinal cord tissue; however, this was not observed. In fact, epidural perfusion cooling reduced gray matter survival. In attempting to understand this finding we observed that the epidural/intraspinal temperature gradient increased as the rate of epidural saline perfusion was increased. This observation has several implications. First, it indicates that we cannot assume that intramedullary temperatures will parallel epidural temperatures. Second, one interpretation of this data is that SCBF is decreased by escalating rates of perfusion epidural cooling. We hypothesize that heat is being lost from the spinal cord to the perfusate and that there is inadequate flow of warm blood to maintain tissue temperature. This is the only paradigm that allows us to understand how the spinal cord could become colder than the epidural fluid. This hypothesis is consistent with the decreased survival of gray matter in the epidurally perfused animals. In other studies of local cooling investigators have shown variations in SCBF, with some reporting a decrease 28,31,76 and others reporting an increase 81 or no change. 56 Systemic hypothermia (27 28 C), however, has been reported to increase SCBF in phenobarbital-anesthetized rats. 57 The potential to achieve hypothermic neuroprotection may be compromised if cooling leads to a 315

9 C. E. Casas, et al. FIG. 8. Graphs demonstrating the effect of epidural fluid perfusion on tissue sparing. A: The mean spared tissue at the epicenter (T-10) was calculated as a preserved/lesion area ratio. B: The mean total preserved spinal cord volumes over the entire lesion region were compared SDs. C and D: The rostral and caudal volumes were then determined based on the epicenter and also compared. E: Longitudinal extent of gray matter loss in micrometers. Values are presented as the means SDs. *p reduction in SCBF because SCBF is already reduced after SCI. 19,20,33,58 Perhaps direct contact of the cool fluid with radicular vessels crossing the epidural space induces vasoconstriction. 22,62 This underexplored variable may serve to explain some of the inconsistencies in the literature regarding the effects of local cooling on post-sci outcome. That reduced gray matter neuroprotection was not correlated with Ba-Be-Br score based outcome indicates the insensitivity of the Ba-Be-Br scoring technique to study thoracic gray matter preservation. The lack of correlation between lesion volumes and gray matter loss may reflect the fact that we set the criteria at three detectable neurons. This quantity of gray matter would represent a very small fraction of the tissue volume. Conclusions Regional epidural perfusion cooling applied 30 minutes after a moderate contusive SCI was ineffective at increasing white matter tissue sparing and caused reduced gray matter neuronal preservation. No significant differences were observed in locomotor recovery over the 6-week time course of the study. Regarding open-field locomotor analysis following a moderate contusive SCI, we showed that FBFVA allows a more detailed examination of complex events, such as forelimb hindlimb coordination, that are difficult to assess using the conventional Ba-Be-Br Scale. Whereas local perfusion cooling has been reported to effectively improve ischemic tolerance of the spinal cord, it appears to be detrimental under the conditions we assessed. Additional studies are required to test the hypothesis that epidural perfusion cooling reduces SCBF. The findings in this investigation validate the reliability of the experimental methods and establish valuable baseline data for subsequent studies of neuroprotection. The data also reinforce the principle that treatments that effectively ameliorate ischemic SCI cannot be assumed to have similar efficacy after contusive spinal cord trauma. We recommend further studies before epidural perfusion cooling is considered for clinical treatment of SCI. References 1. Albin MS, White RJ, Acosta-Rua G, Yashon D: Study of functional recovery produced by delayed localized cooling after spinal cord injury in primates. J Neurosurg 29: , Albin MS, White RJ, Locke GE: Treatment of spinal cord trauma by selective hypothermic perfusion. Surg Forum 16: ,

10 Epidural hypothermic neuroprotection after SCI in rats 3. Allen BT, Davis CG, Osborne D, Karl I: Spinal cord ischemia and reperfusion metabolism: the effect of hypothermia. J Vasc Surg 19: , Basso DM, Beattie MS, Bresnahan JC: A sensitive and reliable locomotor rating scale for open field testing in rats. J Neurotrauma 12:1 21, Basso DM, Beattie MS, Bresnahan JC, Anderson DK, Faden AI, Gruner JA: MASCIS evaluation of open field locomotor scores: effects of experience and teamwork on reliability. Multicenter Animal Spinal Cord Injury Study. J Neurotrauma 13: , Bracken MB, Shepard MJ, Holford TR, Leo-Summers L, Aldrich EF, Fazl M, et al: Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury. Results of the Third National Acute Spinal Cord Injury Randomized Controlled Trial. National Acute Spinal Cord Injury Study. JAMA 277: , Busto R, Dietrich WD, Globus MY, Valdes I, Scheinberg P, Ginsberg MD: Small differences in intraischemic brain temperature critically determine the extent of ischemic neuronal injury. J Cereb Blood Flow Metab 7: , Cambria RP, Davison JK, Carter C, Brewster DC, Chang Y, Clark KA, et al: Epidural cooling for spinal cord protection during thoracoabdominal aneurysm repair: a five-year experience. J Vasc Surg 31: , Cambria RP, Davison JK, Zannetti S, L Italien G, Brewster DC, Gertler JP, et al: Clinical experience with epidural cooling for spinal cord protection during thoracic and thoracoabdominal aneurysm repair. J Vasc Surg 25: , Chatzipanteli K, Yanagawa Y, Marcillo AE, Kraydich S, Yezierski RP, Dietrich WD: Posttraumatic hypothermia reduces polymorphonuclear leukocyte accumulation following spinal cord injury in rats. J Neurotrauma 17: , Colbourne F, Grooms SY, Zukin RS, Buchan AM, Bennett MV: Hypothermia rescues hippocampal CA1 neurons and attenuates down-regulation of the AMPA receptor GluR2 subunit after forebrain ischemia. Proc Natl Acad Sci USA 100: , Colbourne F, Sutherland G, Corbett D: Postischemic hypothermia. A critical appraisal with implications for clinical treatment. Mol Neurobiol 14: , Colon R, Frazier OH, Cooley DA, McAllister HA: Hypothermic regional perfusion for protection of the spinal cord during periods of ischemia. Ann Thorac Surg 43: , Constantini S, Young W: The effects of methylprednisolone and the ganglioside GM1 on acute spinal cord injury in rats. J Neurosurg 80:97 111, Davison JK, Cambria RP, Vierra DJ, Columbia MA, Koustas G: Epidural cooling for regional spinal cord hypothermia during thoracoabdominal aneurysm repair. J Vasc Surg 20: , Dietrich WD, Busto R, Globus MY, Ginsberg MD: Brain damage and temperature: cellular and molecular mechanisms. Adv Neurol 71: , Dimar JR II, Shields CB, Zhang YP, Burke DA, Raque GH, Glassman SD: The role of directly applied hypothermia in spinal cord injury. Spine 25: , Ducker TB, Hamit HF: Experimental treatments of acute spinal cord injury. J Neurosurg 30: , Ducker TB, Salcman M, Lucas JT, Garrison WB, Perot PL Jr: Experimental spinal cord trauma, II: Blood flow, tissue oxygen, evoked potentials in both paretic and plegic monkeys. Surg Neurol 10:64 70, Ducker TB, Salcman M, Perot PL Jr, Ballantine D: Experimental spinal cord trauma, I: Correlation of blood flow, tissue oxygen and neurologic status in the dog. Surg Neurol 10: 60 63, Frank SM, Beattie C, Christopherson R, Norris EJ, Perler BA, Williams BM: Unintentional hypothermia is associated with postoperative myocardial ischemia. The Perioperative Ischemia Randomized Anesthesia Trial Study Group. Anesthesiology 78: , Gardner CA, Webb RC: Cold-induced vasodilatation in isolated, perfused rat tail artery. Am J Physiol 251:H176 H181, Green B, Khan T, Raimondi AJ: Local hypothermia as treatment of experimentally induced spinal cord contusion: quantitative analysis of beneficient effect. Surg Forum 24: , Gruner JA: A monitored contusion model of spinal cord injury in the rat. J Neurotrauma 9: , Guest J, Vanni S, Silbert L: Mild hypothermia, blood loss and complications in elective spinal surgery. Spine J 4: , Guha A, Tator CH: Acute cardiovascular effects of experimental spinal cord injury. J Trauma 28: , Hansebout RR, Kuchner EF, Romero-Sierra C: Effects of local hypothermia and of steroids upon recovery from experimental spinal cord compression injury. Surg Neurol 4: , Hansebout RR, Lamont RN, Kamath MV: The effects of local cooling on canine spinal cord blood flow. Can J Neurol Sci 12: 83 87, Inamasu J, Nakamura Y, Ichikizaki K: Induced hypothermia in experimental traumatic spinal cord injury: an update. J Neurol Sci 209:55 60, Kader A, Frazzini VI, Baker CJ, Soloman RA, Trifiletti RR: Effect of mild hypothermia on nitric oxide synthesis during focal cerebral ischemia. Neurosurgery 35: , Kano T, Sadanaga M, Sakamoto M, Higashi K, Matsumoto M: Effects of systemic cooling and rewarming on the evoked spinal cord potentials and local spinal cord blood flow in dogs. Anesth Analg 78: , Karabiyikoglu M, Han HS, Yenari MA, Steinberg GK: Attenuation of nitric oxide synthase isoform expression by mild hypothermia after focal cerebral ischemia: variations depending on timing of cooling. J Neurosurg 98: , Kelly DL Jr, Lassiter KR, Calogero JA, Alexander E Jr: Effects of local hypothermia and tissue oxygen studies in experimental paraplegia. J Neurosurg 33: , Koizumi H, Povlishock JT: Posttraumatic hypothermia in the treatment of axonal damage in an animal model of traumatic axonal injury. J Neurosurg 89: , Kuchner EF, Hansebout RR: Combined steroid and hypothermia treatment of experimental spinal cord injury. Surg Neurol 6: , Kuchner EF, Hansebout RR, Pappius HM: Effects of dexamethasone and of local hypothermia on early and late tissue electrolyte changes in experimental spinal cord injury. J Spinal Disord 13: , Kurz A, Sessler DI, Lenhardt R: Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. Study of Wound Infection and Temperature Group. N Engl J Med 334: , Laptook AR, Corbett RJ, Burns D, Sterett R: Neonatal ischemic neuroprotection by modest hypothermia is associated with attenuated brain acidosis. Stroke 26: , Mackenzie CF, Shin B, Krishnaprasad D, McCormack F, Illingworth W: Assessment of cardiac and respiratory function during surgery on patients with acute quadriplegia. J Neurosurg 62: , Malatova Z, Vanicky I, Galik J, Marsala M: Epidural perfusion cooling protects against spinal cord ischemia in rabbits. An evaluation of cholinergic function. Mol Chem Neuropathol 25: 81 96, Marion DW, Leonov Y, Ginsberg M, Katz LM, Kochanek PM, Lechleuthner A, et al: Resuscitative hypothermia. Crit Care Med 24:S81 S89, Marion DW, Penrod LE, Kelsey SF, Obrist WD, Kochanek PM, Palmer AM, et al: Treatment of traumatic brain injury with moderate hypothermia. N Engl J Med 336: , Marion DW, White MJ: Treatment of experimental brain injury with moderate hypothermia and 21-aminosteroids. J Neurotrauma 13: ,

11 C. E. Casas, et al. 44. Marsala M, Galik J, Ishikawa T, Yaksh TL: Technique of selective spinal cord cooling in rat: methodology and application. J Neurosci Methods 74:97 106, Marsala M, Vanicky I, Galik J, Radonak J, Kundrat I, Marsala J: Panmyelic epidural cooling protects against ischemic spinal cord damage. J Surg Res 55:21 31, Marsala M, Vanicky I, Yaksh TL: Effect of graded hypothermia (27 to 34 C) on behavioral function, histopathology, and spinal blood flow after spinal ischemia in rat. Stroke 25: , Martelli E, Cho JS, Mozes G, Gloviczki P: Epidural cooling for the prevention of ischemic injury to the spinal cord during aortic occlusion in a rabbit model: determination of the optimal temperature. J Vasc Surg 35: , Martinez-Arizala A, Green BA: Hypothermia in spinal cord injury. J Neurotrauma 9 (Suppl 2):S497 S505, Molano Mdel R, Broton JG, Bean JA, Calancie B: Complications associated with the prophylactic use of methylprednisolone during surgical stabilization after spinal cord injury. J Neurosurg 96: , Pontius RG, Brockman HL, Hardy EG, Cooley DA, DeBakey ME: The use of hypothermia in the prevention of paraplegia following temporary aortic occlusion: experimental observations. Surgery 36:33 38, Quinones-Hinojosa A, Malek JY, Ames A III, Ogilvy CS, Maynard KI: Metabolic effects of hypothermia and its neuroprotective effects on the recovery of metabolic and electrophysiological function in the ischemic retina in vitro. Neurosurgery 52: , Rittenhouse EA, Mohri H, Morgan BC, Dillard DH, Merendino KA: Electrocardiographic changes in infants undergoing surface-induced deep hypothermia for open-heart surgery. Am Heart J 79: , Robertson CS, Foltz R, Grossman RG, Goodman JC: Protection against experimental ischemic spinal cord injury. J Neurosurg 64: , Rokkas CK, Sundaresan S, Shuman TA, Palazzo RS, Nitta T, Depotis GJ, et al: Profound systemic hypothermia protects the spinal cord in a primate model of spinal cord ischemia. J Thorac Cardiovasc Surg 106: , Rosomoff HL, Holaday DA: Cerebral blood flow and cerebral oxygen consumption during hypothermia. Am J Physiol 179: 85 88, Sakamoto T, Iwai A, Monafo WW: Regional blood flow in transected rat spinal cord during hypothermia. Am J Physiol 259:H1649 H1654, Sakamoto T, Monafo WW: Regional blood flow in the brain and spinal cord of hypothermic rats. Am J Physiol 257: H785 H790, Sandler AN, Tator CH: Effect of acute spinal cord compression injury on regional spinal cord blood flow in primates. J Neurosurg 45: , Sands MP, Sato S, Mohri H, Guntheroth WG, Merendino KA: Electrocardiographic changes during surface-induced deep hypothermia. The influence of ether, halothane, carbon dioxide, and perfusion rewarming. Ann Thorac Surg 19: , Schaller B, Graf R: Hypothermia and stroke: the pathophysiological background. Pathophysiology 10:7 35, Sessler DI: Deliberate mild hypothermia. J Neurosurg Anesthesiol 7:38 46, Shepherd JT, Rusch NJ, Vanhoutte PM: Effect of cold on the blood vessel wall. Gen Pharmacol 14:61 64, Tabayashi K, Niibori K, Konno H, Mohri H: Protection from postischemic spinal cord injury by perfusion cooling of the epidural space. Ann Thorac Surg 56: , Takeda Y, Namba K, Higuchi T, Hagioka S, Takata K, Hirakawa M, et al: Quantitative evaluation of the neuroprotective effects of hypothermia ranging from 34 C to 31 C on brain ischemia in gerbils and determination of the mechanism of neuroprotection. Crit Care Med 31: , Tator CH, Deecke L: Studies of the treatment and pathophysiology of acute spinal cord injury in primates. Paraplegia 10: , Tator CH, Deecke L: Value of normothermic perfusion, hypothermic perfusion, and durotomy in the treatment of experimental acute spinal cord trauma. J Neurosurg 39:52 64, Thomas R, Hessel EA II, Harker LA, Sands MP, Dillard DH: Platelet function during and after deep surface hypothermia. J Surg Res 31: , Tibbs PA, Young B, Todd EP, McAllister RG Jr, Hubbard S: Studies of experimental cervical spinal cord transection. Part IV: Effects of cervical spinal cord transection on myocardial blood flow in anesthetized dogs. J Neurosurg 52: , Troll GF, Dohrmann GJ: Anaesthesia of the spinal cord-injured patient: cardiovascular problems and their management. Paraplegia 13: , Tuzgen S, Kaynar MY, Guner A, Gumustas K, Belce A, Etus V, et al: The effect of epidural cooling on lipid peroxidation after experimental spinal cord injury. Spinal Cord 36: , Tymianski M, Sattler R, Zabramski JM, Spetzler RF: Characterization of neuroprotection from excitotoxicity by moderate and profound hypothermia in cultured cortical neurons unmasks a temperature-insensitive component of glutamate neurotoxicity. J Cereb Blood Flow Metab 18: , Uihlein A, MacCarty CS, Michenfelder JD, Terry HR Jr, Daw EF: Deep hypothermia and surgical treatment of intracranial aneurysms. A five-year survey. JAMA 195: , Vacanti FX, Ames A III: Mild hypothermia and Mg++ protect against irreversible damage during CNS ischemia. Stroke 15: , Vanicky I, Marsala M, Galik J, Marsala J: Epidural perfusion cooling protection against protracted spinal cord ischemia in rabbits. J Neurosurg 79: , Wakamatsu H, Matsumoto M, Nakakimura K, Sakabe T: The effects of moderate hypothermia and intrathecal tetracaine on glutamate concentrations of intrathecal dialysate and neurologic and histopathologic outcome in transient spinal cord ischemia in rabbits. Anesth Analg 88:56 62, Westergren H, Farooque M, Olsson Y, Holtz A: Spinal cord blood flow changes following systemic hypothermia and spinal cord compression injury: an experimental study in the rat using Laser-Doppler flowmetry. Spinal Cord 39:74 84, Williams GD, Dardzinski BJ, Buckalew AR, Smith MB: Modest hypothermia preserves cerebral energy metabolism during hypoxia-ischemia and correlates with brain damage: a 31P nuclear magnetic resonance study in unanesthetized neonatal rats. Pediatr Res 42: , Winfree CH, Baker KZ, Connolly ES: Perioperative normothermia and surgical-wound infection. N Engl J Med 335: , Young W: Spinal cord contusion models. Prog Brain Res 137: , Yu CG, Jimenez O, Marcillo AE, Weider B, Bangerter K, Dietrich WD, et al: Beneficial effects of modest systemic hypothermia on locomotor function and histopathological damage following contusion-induced spinal cord injury in rats. J Neurosurg 93:85 93, Zielonka JS, Wagner FC Jr, Dohrmann GJ: Alterations in spinal cord blood flow during local hypothermia. Surg Forum 25: , Zornow MH: Inhibition of glutamate release: a possible mechanism of hypothermic neuroprotection. J Neurosurg Anesthesiol 7: , 1995 Manuscript received April 20, Accepted in final form October 22, Address reprint requests to: James D. Guest, M.D., Ph.D., Department of Neurological Surgery, University of Miami, Lois Pope LIFE Center, 1095 NW 14th Terrace (D4-6), Miami, Florida jguest@med.miami.edu. 318