brain research 1618 (2015) 55–60
Available online at www.sciencedirect.com
www.elsevier.com/locate/brainres
Research Report
Clinical indicators of paraplegia underplay universal spinal cord neuronal injury from transient aortic occlusion$ Marshall T Bella,n, Ferenc Puskasb, Daine T Bennetta, Joseph C. Cleveland Jra, Paco S. Hersonb, Joshua M. Maresa, Xainzhong Menga,b, Michael J. Weyanta, David A. Fullertona, T. Brett Reecea a
Department of Surgery, University of Colorado, Denver, CO, USA Department of Anesthesiology, University of Colorado, Denver, CO, USA
b
art i cle i nfo
ab st rac t
Article history:
Paraplegia following complex aortic intervention relies on crude evaluation of lower
Accepted 28 April 2015
extremity strength such as whether the patient can lift their legs or flex the ankle. Little
Available online 22 May 2015
attention has been given to the possible long-term neurologic sequelae following these
Keywords:
procedures in patients appearing functionally normal. We hypothesize that mice subjected
Aorta
to minimal ischemic time will have functional and histological changes despite the gross
Spinal cord ischemia–reperfusion
appearance of normal function.
Neurologic injury
Male mice underwent 3 min of aortic occlusion (n ¼14) or sham surgery (n ¼4) via a
Spinal cord
median sternotomy. Neurologic function was graded by Basso Motor Score (BMS) pre-
Paraplegia
operatively and at 24 h intervals after reperfusion. Mice appearing functionally normal and sham mice were placed on a walking beam and recorded on high-definition, for singleframe motion analysis. After 96hrs, spinal cords were removed for histological analysis. Following 3 min of ischemia, functional outcomes were split evenly with either mice displaying almost normal function n ¼7 or near complete paraplegia n ¼7. Additionally, single-frame motion analysis revealed significant changes in gait. Histologically, there was a significant stepwise reduction of neuronal viability, with even the normal function ischemic group demonstrating significant loss of neurons. Despite the appearance of normal function, temporary ischemia induced marked cytoarchitectural changes and neuronal degeneration. Furthermore high-definition gait analysis revealed significant changes in gait and activity following thoracic aortic occlusion. These data suggest that all patients undergoing procedures, even with short ischemic times, may have spinal cord injury that is not evident clinically. & 2015 Elsevier B.V. All rights reserved.
☆
Southern Thoracic Surgical Association 60th Annual Meeting, November 1st, 2013, Scottsdale, AZ. Corresponding author at: School of Medicine, Division of Cardiothoracic Surgery, University of Colorado, 12631 East 17th Avenue, Aurora 80045, CO, USA. Fax: þ1 303 724 2806. E-mail address: [email protected] (M. Bell). n
http://dx.doi.org/10.1016/j.brainres.2015.04.053 0006-8993/& 2015 Elsevier B.V. All rights reserved.
56
1.
brain research 1618 (2015) 55–60
overt paralysis did have a significant reduction in rump height indices when compared to sham mice (Table 2).
Introduction
Paraplegia following complex aortic intervention remains a devastating complication (Becker et al., 2013). Evaluation of spinal cord injury following aortic surgery has long been thought to be either complete motor deficit or complete absence of injury. Other organ systems susceptible to ischemia, such as the kidneys, have demonstrated that a critical mass of cellular injury is required for organ dysfunction to be apparent (Simmons et al., 2011, 2012). Previous studies of spinal cord hypoxia/ischemia have made it clear that the spinal cord neuronal injury is not an all or nothing phenomenon with injured neurons in almost all ischemic groups. However, many of these groups had relatively normal motor function despite the clear neuronal injury (Ozkisacik et al., 2011; Smith et al., 2013). The aim of this study was to not only demonstrate that spinal cord motor neuron injury can occur in the absence of gross functional deficits, but that the evaluation of this injury may be inadequate. We hypothesize that mice subjected to minimal ischemic time will have functional and histological changes despite the gross appearance of normal function.
2.
Results
2.1.
Qualitative functional assessment
2.3.
Histological
All Mice that underwent thoracic aortic occlusion had significant cyto-architectural decay with increased vacuolization, pyknotic nuclei and a significant disruption of the greywhite matter junction. Though these findings were more pronounced in paraplegic mice, they were ubiquitous in all mice following thoracic aortic occlusion (Fig. 2).
2.4.
Neuronal viability
Following sham surgery or thoracic aortic occlusion there was a step wise reduction neuronal viability. In mice subjected to IR surgery without flaccid paralysis, a significant reduction in both total anterior horn neurons (A) and lamina IX alpha-motor neurons (B) was observed. Mice with flaccid paralysis had an additional reduction in the number anterior horn motor neurons that was significant less than both other groups (Fig. 3).
3.
Discussion
The current data demonstrate several telling points with regard to the evaluation of motor function following reperfu-
Sham mice (n ¼4) had no observable functional deficits preoperatively and throughout duration of the experiment. There was no significant differences in experimental variables (weight, temperature or minimal perfusion) between mice that develop flaccid paralysis and those that were functionally normal (Table 1). Mice subjected to thoracic aortic occlusion for 3 min developed either overt paralysis (n ¼7) by 48 h or displayed minimal hind-limb deficits that were not statistically different from sham mice (n¼ 7) (Fig. 1).
2.2.
Quantitative functional assessment
While functional analysis using basso mouse score was unable to detected differences from sham mice and mice not displaying overt paralysis, single-frame motion analysis was able to detect significant differences. Assessment of hind-limb strength was performed by measuring the rump height indices of sham mice and mice that underwent IR surgery but did not exhibit gross functional deficits. Mice subjected to thoracic aortic occlusion that did not display
Fig. 1 – Functional outcomes following thoracic aortic occlusion or sham surgery. Mice that underwent thoracic aortic occlusion for three minutes had either minimal hindlimb deficits that were not statistically significant from sham mice or developed severe hind-limb deficits that were statistically significant from all other groups (*p-value o0.05).
Table 1 – Operative variables are highlighted above. Perioperative weight, in addition to average intraoperative perfusion and temperature were recorded ischemic groups. Weight was recorded preoperatively while temperature and perfusion were recorded at thirty second intervals.
Weight (gm) Average temp (1C) Minimal perfusion units (PU)
IR no paralysis
IR paralysis
p Value
27.873.1 36.47.33 57.8724.7
27.071.3 36.47.26 63.0715.8
0.54 0.97 0.64
brain research 1618 (2015) 55–60
sion of spinal cord ischemia. First, spinal cord motor neuron injury is universal with temporary aortic occlusion. Second, even the nine point Basso motor score is not sensitive enough to detect the subtle differences in neurological function resulting from ischemia and reperfusion injury. These points suggest that both diagnosis and treatment of spinal cord injury is severely lacking in patients undergoing aortic intervention. In order to further delineate spinal cord ischemia–reperfusion injury, we enlisted a murine model that reproduces the clinical entities of both immediate and delayed paraplegia seen following clinically thoracic aortic occlusion (Smith et al., 2011). This study, among others, demonstrated motor neuron degeneration in all mice undergoing ischemia and reperfusion independent of the obvious motor functional preservation (Smith et al., 2013) Thus, we began to investigate the effects of a minimal ischemic time on functional outcomes. Based on our previous studies 48 min of thoracic aortic occlusion resulted in immediate paraplegia, 4 min resulted in a delayed paraplegia and 3 min resulted in a minimal hind-limb deficit. However, the binary nature or injury, described as paraplegia or not, does not tell the whole story. When an ischemic time of 3 min was further investigated with a larger sample size we found mice exhibited a bimodal distribution of injury. After 48 h of reperfusion, mice had Table 2 – Rump height indices (RHI) were compared in sham mice and mice that did not progress to paralysis. Pre-operatively there was no difference in RHI. Following surgery, there was a significant reduction in RHI in mice that underwent thoracic aortic occlusion when compared to sham mice at all time points post-operatively. indicating a reduction in hind-limb strength. Rump Height Index
Preop 24 hrs 48 hrs 72 hrs 96 hrs
Sham
IR
P value
0.857.01 0.867.02 0.837.01 0.847.01 0.837.01
0.847.02 0.727.03 0.717.04 0.677.03 0.617.04
0.72 0.01 0.06 0.01 0.01
57
complete paralysis or minimal hind-limb deficits. Importantly, these spinal cord differences were observed in the absence of significant procedural variability. Variables including temperature, quantity of distal perfusion, and ischemic time were identical despite different functional outcomes. While the differences between the bimodal functional outcomes with similar ischemic insults deserve additional attention in future study, the current study focuses on the spinal cord changes in the functionally intact ischemic animals We next pursued the extent of histological changes in our 3 groups with a particular interest on mice with intact hindlimb function despite thoracic aortic occlusion. Mice that had been subjected to ischemia and reperfusion but did not exhibit gross functional deficits were able to ambulate despite a 40% reduction in motor neurons. Mice with a 70% reduction in motor neurons exhibited near complete function deficits. This finding suggests that once neuronal decay has extended beyond a critical mass, paraplegia will occur. Currently, only after dysfunction beyond this scope will be treated in most postoperative algorithms. In most programs, this would be limited to hypertension and CSF drainage rather than preventative measures. The current data The evolution of surgical adjuncts such as hypothermia, lumbar drains, and intercostal artery reimplantation has reduced the risk of paraplegia following aortic interventions (Okita, 2011). Unfortunately, up to 20% of patients undergoing high-risk repairs are still affected (Conrad et al., 2008; Kouchoukos et al., 2013). Most treatment algorithms for spinal cord injury are limited to extreme risk patients or patients exhibiting injury. Despite potential use of left heart bypass and hypothermia, the degree of motor neuron loss demonstrated in this study suggests the full treatment is reserved only for the worst injuries. Further treatment paradigms need to be developed for prevention in all patients at risk for several reasons. Clearly, further reduction of injury would limit exposure to a reduced life-expectancy and lifetime of debility experienced by paraplegic patients (Messé et al., 2008). The extent of injury that goes clinically undetected is unknown. More subtle injury may affect motor function over time. The relationship between functional decline and neuronal injury has been most widely studied in neurodegenerative
Fig. 2 – Representative images of anterior horns of mice in all three experimental groups (Magnification 20 ). All mice that underwent thoracic aortic occlusion had significant cyto-architectural disruption, increased vacuolization and a reduction in normal appearing anterior horn neurons. While these finding were present in all mice subjected to thoracic aortic occlusion they were much more pronounced in mice that became paraplegic. Black arrows depict injured anterior horn motor neurons.
58
brain research 1618 (2015) 55–60
Fig. 3 – Normal appearing neurons were quantified in both the anterior horn (A) and Lamina IX motor neurons (B) with representative numerical data (C). There was a significant reduction in normal appearing neurons in both groups that underwent thoracic aortic occlusion when compared to sham mice (*po0.05) Among mice that underwent thoracic aortic occlusion there was an additional reduction in anterior horn neurons in mice that became paraplegic. (þþpo0.05).
models in which clear correlation between neuronal injury and function can be defined (Golberg et al., 2011; Wang et al., 2012). Additionally, traumatic brain injury models have shown that minimal injury, not associated with functional deficits, the ischemic injury results in activation neurodegenerative cascades (Tashlykov et al., 2009). While many similarities exist between the brain and the spinal cord, the spinal cord is physiologically and functionally different than the brain. Spinal cord motor neurons represent a small population of spinal cord neurons (Bjugn, 1993) and are the largest of multipolar cells within the central nervous system (Geinismann, 1971). Additionally, motor neurons in spinal cord are much more susceptible to oxidative stress and mitochondrial dysfunction than cortical motor neurons (Panov et al., 2011). While spinal cord neuronal injury has been studied in perinatal rodents during hypoxic stress (Clancy et al., 1989), it was not been widely investigated following transient spinal cord ischemia. Only with further identification of the more subtle injury can the implications of this injury be identified. Therefore, spinal cord motor neurons may even have more detrimental long-term effects of this potentially unapparent injury. Further, we believe this model is representative of both open and endovascular treatments of complex aortic pathology. The rates of delayed paraplegia remain very similar between this approaches and both are thought to be ischemia and reperfusion in nature, with the cross clamp, clearly the ischemic induction in open cases while the creation of a watershed zone of perfusion from the excluded zone of the thoracic aorta (Eagleton and Greenberg, 2010). The perfusion to this watershed zone should improve over time, but according to Griepps group this may take up to 120 h (Etz et al., 2010). Supporting these neurons through this potential malperfusion should be inherent. However, our current treatment paradigms continue to treat the symptoms of motor neurons dysfunction rather than a protective milieu for the at risk neurons. The current study suggests that better strategies for metabolic protection of these neurons are not only necessary, but also should be applied to any patient undergoing complex aortic intervention. As treatment of interstitial and familial aortopathy presses for earlier intervention (Samadi et al., 2012), more patients will have longer post-
intervention lives with which to live with potential injury. The less pervasive deficits are not supported by current algorithm, so it could be argued that most any aortic intervention could benefit from preventative therapies. This study, however, is not without its limitations. The model has been criticized for the extent of injury relative to the short ischemic time. While the ischemic time is short relative to clinical ischemia, the metabolic rate of mice is much higher than humans (Kleiber, 1975). Further, there is no adjunctive protection from the ischemia in the model such as distal perfusion resulting in a higher ischemic burden for the model. In addition, we have followed the model out for up to 14 days without long term changes in neurological function. Despite the appearance of normal function, IR surgery induced marked cyto-architectural changes and neuronal degeneration. Furthermore high-definition gait analysis revealed significant changes in gait and activity following thoracic aortic occlusion. These data suggest that all patients undergoing surgery, even with short ischemic times, may have spinal cord injury that is not evident clinically. In summary, spinal cord injury may be ubiquitous in aortic intervention necessitating the development of neuroprotective therapies for all these procedures instead of limiting them to patients at high risk for injury or demonstrating gross deficits by exam.
4.
Experimental procedures
4.1.
Animal procedures
The Animal Care and Use Committee at the University of Colorado at Denver Health Sciences Center approved all experiments, and this investigation conformed to the Guide for the Care and Use of Laboratory Animals published by the National Institute of Health (www.nap.edu/catalog/5140.html). Adult, male C57BL/6mice were used for all experimentations.
4.2. Aortic cross clamping: Ischemia–reperfusion (IR) surgery Mice were anesthetized using 2% isoflurane and placed in the supine position. Surgery was performed under normothermic
brain research 1618 (2015) 55–60
conditions with core body temperature maintained at 36.570.5 1C using a rectal temperature probe and automatic temperature adjusting bed (Vestavia Scientific, Birmingham, AL). The aortic arch was exposed using a cervicothoracic approach as previously described. A disruption arterial flow to the spinal cord was achieved by placing vascular clamps on the aortic arch distal to the left common carotid artery and the subclavian artery for 3 min. A laser doppler blood flow monitor (Moor Instruments, UK) measuring femoral artery flow in perfusion units (PU) was placed over the left femoral artery. Prior to aortic/subclavian occlusion max perfusion of greater than 900 PU was documented. A 90% reduction in distal flow (less than 90 PU) during aortic occlusion was obtained in all mice used in ischemic groups. If mice did not have a document decreased in distal flow of more than 90% they were humanely euthanized and not included in functional outcomes.
4.3.
Qualitative functional assessment
The Basso Mouse Scale for locomotion (Basso et al., 2006) which ranges from a score of 0 for complete paraplegia to a score of 9 for normal function was used to quantify hind-limb function in mice after ischemia. Functional scores were recorded preoperatively and at 24, 48, 72 and 96 h following surgery.
4.4.
Quantitative functional assessment
Mice without overt functional deficits (Sham and IR) were place on a walking beam apparatus to analyze gait are previously described (Semler et al., 2011). Images were recorded by a high-definition camera (Panasonic HC-V500M, Newark, NJ). Single-frame images were analyzed using HD Writer AE 4.0 (Panasonic, Newark, NJ) and processed by Image Tool 3.0 (UT Health Science Center, San Antonio, TX). Rump height indices were used for calculation of hind-limb strength. This calculation was obtained by dividing the recorded height of the base of the tail above the surface of the beam by the width recorded of the beam. Three individual measurements were taking at 24 h apart by a blinded observer and averaged.
4.5.
Histological analysis
Ninety-six hrs after surgery, the animals were sacrificed. The vertebral column was removed en bloc fromT8-L3. Spinal cords were removed from the vertebral column by injection of phosphate buffered saline (PBS, pH 7.4) into the spinal column. Spinal cords were then transferred to 4% formalin where they remained for at least 24 h prior to paraffin embedding, sectioning, and hematoxylin and eosin staining. Sectioning and staining was performed by a blinded researcher who provided cranial and caudal representations of each specimen.
4.6.
Quantification of neuronal injury
Sections were evaluated by a blindly observer for neuronal viability and quantified as neurons per anterior horn. Neurons
59
within the anterior horn that contained prominent nucleoli and loose chromatin were considered normal (Kurita et al., 2005). Quantification was performed for both total anterior horn neurons and motor neurons in Lamina IX neurons.
4.7.
Statistical analysis
Statistical analysis was performed using repeated measures and the Kruskal-Wallis nonparametric ANOVA with Bonferroni/Dunn post-hoc test using, StatView (SAS Institute Inc., Cary, CA). Data are presented as mean7standard error. A pvalue o0.05 was considered significant for all statistical comparisons.
r e f e r e nc e s
Basso, D.M., Fisher, L.C., Anderson, A.J., Jakeman, L.B., McTigue, D. M., Popovich, P.G., 2006. Basso Mouse Scale for locomotion detects differences in recovery after spinal cord injury in five common mouse strains. J. Neurotrauma 23, 635–659. Becker, D.A., McGarvey, M.L., Rojvirat, C., Bavaria, J.E., Messe´, S.R., 2013. Predictors of outcome in patients with spinal cord ischemia after open aortic repair. Neurocrit. Care 18, 70–74. Bjugn, R., 1993. The use of the optical disector to estimate the number of neurons, glial and endothelial cells in the spinal cord of the mouse—with a comparative note on the rat spinal cord. Brain Res. 627, 25–33. Clancy, R.R., Sladky, J.T., Rorke, L.B., 1989. Hypoxic–ischemic spinal cord injury following perinatal asphyxia. Ann. Neurol. 25, 185–189. Conrad, M.F., Crawford, R.S., Davison, J.K., Cambria, R.P., 2008. Thoracoabdominal aneurysm repair: a 20-year perspective S856-S861. Ann. Thorac. Surg. 8. Eagleton, M.J., Greenberg, R.K., 2010. Spinal and visceral ischemia after endovascular aortic repair. J. Cardiovasc. Surg. 51, 71–83. Etz, C.D., Zoli, S., Bischoff, M.S., Bodian, C., Di Luozzo, G., Griepp, R.B., 2010. Measuring the collateral network pressure to minimize paraplegia risk in thoracoabdominal aneurysm resection. J. Thorac. Cardiovasc. Surg. 140, S125–S130. Geinismann, Y.Y., 1971. Determination of nerve and glial cell volumes in histological section. Brain Res. 26, 235–246. Golberg, N.R., Haack, A.K., Lims, N.S., Janson, O.K., Meshul, C.K., 2011. Dopaminergic and beharvioral corraletes of progressive lesioning of the nigrostriatal pathway with 1-methyl-4phenyl-1,2,3,6 tetrahydropyridine. Neuroscience 180, 256–271. Kleiber, M., 1975. The metabolic rate per unit body weight. J. Theor. Biol. 53, 199–204. Kouchoukos, N.T., Kulik, A., Castner, C.F., 2013. Outcomes after thoracoabdominal aortic aneurysm repair using hypothermic circulatory arrest. J. Thorac. Cardiovas. Surg 145, S139–S141. Kurita, N., Kawaguchi, M., Horiuchi, T., Inoue, S., Sakamoto, T., Nakamura, M., Konishi, N., Furuya, H., 2005. An evaluation of white matter injury after spinal cord ischemia in rats: a comparison with gray matter injury. Anesth. Analg. 100, 847–854. Messe´, S.R., Bavaria, J.E., Mullen, M., Cheung, A.T., Augoustides, J. G., Futsche, J., Woo, E.Y., Szeto, W.Y., Pochettino, A., Woo, Y.J., Kasner, S.E., McCarvey, M., 2008. Neurologic outcomes from high risk descending thoracic and thoracoabdominal aortic operations in the era of endovascular repair. Neurocrit. Care 9, 344–351. Okita, Y., 2011. Fighting spinal cord complication during surgery for thoracoabdominal aortic disease. Gen. Thorac. Cardiovasc. Surg. 59, 79–90.
60
brain research 1618 (2015) 55–60
Ozkisacik, E.A., Discigil, B., Boga, M., Grucun, U., Badak, M.I., Kurtoglu, T., Dikicioglu, E., 2011. Effects of cyclosporin a on neurological outcome nd serum biomarkers in the same setting of spinal cord ischemia model. Ann. Vasc. Surg. 20, 243–249. Panov, A.V., Kubalik, N., Zinchenko, N., Ridings, D.M., Radoff, D.A., Hemendinger, R., Brooks, B.R., Bonkosvsky, H.L., 2011. Metabolic and functional differences between brain and spinal cord mitochondria underlie different predisposition to pathology. Am. J. Physiol. Regul. Integr. Comp. Physiol 300, R844–R854. Samadi, A., Detaint, D., Roy, C., Arnoult, F., Delorme, G., Gautier, M., Milleron, O., Raoux, F., Meuleman, C., Hvass, U., Hamroun, D., Beroud, C., Tubach, R., Boileau, C., Jondeau, G., 2012. Surgical management of patients with Marfan syndrome: evolution throughout the years. Arch. Cardiovasc. Dis. 105, 84–90. Simmons, M.N., Fergany, A.F., Campbell, S.C., 2011. Effect of parenchymal volume preservation on kidney function after partial nephrectomy. J. Urol. 186, 405–410. Simmons, M.N., Hillyer, S.P., Lee, B.H., Fergany, A.F., Kaouk, J., Campbell, S.C., 2012. Functional recovery after partial nephrectomy: effects of volume loss and ischemic injury. J. Urol. 187, 1667–1673.
Semler, J., Wellmann, K., Wirth, F., Stein, G., Angelova, S., Ashrafi, M., Schempf, G., Ankerne, J., Ozsoy, O., Ozsoy, U., Angelov, D. N., Irintchev, A., 2011. Objective measures of motor dysfunction after compression spinal cord injury in adult rats: correlations with locomotor rating scores. J. Neurotrauma 28, 1247–1258. Smith, P.D., Puskas, F., Meng, X., Cho, D., Cleveland, J.C., Jr., Weyant, M.J., Fullerton, D.A., Reece, T.B., 2011. Ischemic dose– response in the spinal cord: both immediate and delayed paraplegia. J. Surg. Res. 7, 1–7. Smith, P.D., Bell, M.T., Puskas, F., Meng, X., Cleveland Jr., J.C., Weyant, M.J., Fullerton, D.A., Reece, T.B., 2013. Preservation of motor function after spinal cord ischemia and reperfusion injury through microglial inhibition. Ann. Thorac. Surg. 95, 1647–1653. Tashlykov, V., Katz, Y., Volkov, A., Gazit, V., Schreiber, S., Zohar, O., Pick, C.G., 2009. Minimal traumatic brain injury induced apoptotic cell death in mice. J. Mol. Neurosci. 37, 16–24. Wang, X.H., Lu, G., Hu, X., Tsang, K.S., Kwong, W.H., Wu, F.X., Meng, H.W., Jian, S., Liu, S.W., Ng, H.K., Poon, W.S., 2012. Progressive assesment of gait and neurochemial correlation in a classical murine model of Parkinson’s disease. BMC Neruosci 13, 142.
Available online at www.sciencedirect.com
www.elsevier.com/locate/brainres
Research Report
Clinical indicators of paraplegia underplay universal spinal cord neuronal injury from transient aortic occlusion$ Marshall T Bella,n, Ferenc Puskasb, Daine T Bennetta, Joseph C. Cleveland Jra, Paco S. Hersonb, Joshua M. Maresa, Xainzhong Menga,b, Michael J. Weyanta, David A. Fullertona, T. Brett Reecea a
Department of Surgery, University of Colorado, Denver, CO, USA Department of Anesthesiology, University of Colorado, Denver, CO, USA
b
art i cle i nfo
ab st rac t
Article history:
Paraplegia following complex aortic intervention relies on crude evaluation of lower
Accepted 28 April 2015
extremity strength such as whether the patient can lift their legs or flex the ankle. Little
Available online 22 May 2015
attention has been given to the possible long-term neurologic sequelae following these
Keywords:
procedures in patients appearing functionally normal. We hypothesize that mice subjected
Aorta
to minimal ischemic time will have functional and histological changes despite the gross
Spinal cord ischemia–reperfusion
appearance of normal function.
Neurologic injury
Male mice underwent 3 min of aortic occlusion (n ¼14) or sham surgery (n ¼4) via a
Spinal cord
median sternotomy. Neurologic function was graded by Basso Motor Score (BMS) pre-
Paraplegia
operatively and at 24 h intervals after reperfusion. Mice appearing functionally normal and sham mice were placed on a walking beam and recorded on high-definition, for singleframe motion analysis. After 96hrs, spinal cords were removed for histological analysis. Following 3 min of ischemia, functional outcomes were split evenly with either mice displaying almost normal function n ¼7 or near complete paraplegia n ¼7. Additionally, single-frame motion analysis revealed significant changes in gait. Histologically, there was a significant stepwise reduction of neuronal viability, with even the normal function ischemic group demonstrating significant loss of neurons. Despite the appearance of normal function, temporary ischemia induced marked cytoarchitectural changes and neuronal degeneration. Furthermore high-definition gait analysis revealed significant changes in gait and activity following thoracic aortic occlusion. These data suggest that all patients undergoing procedures, even with short ischemic times, may have spinal cord injury that is not evident clinically. & 2015 Elsevier B.V. All rights reserved.
☆
Southern Thoracic Surgical Association 60th Annual Meeting, November 1st, 2013, Scottsdale, AZ. Corresponding author at: School of Medicine, Division of Cardiothoracic Surgery, University of Colorado, 12631 East 17th Avenue, Aurora 80045, CO, USA. Fax: þ1 303 724 2806. E-mail address: [email protected] (M. Bell). n
http://dx.doi.org/10.1016/j.brainres.2015.04.053 0006-8993/& 2015 Elsevier B.V. All rights reserved.
56
1.
brain research 1618 (2015) 55–60
overt paralysis did have a significant reduction in rump height indices when compared to sham mice (Table 2).
Introduction
Paraplegia following complex aortic intervention remains a devastating complication (Becker et al., 2013). Evaluation of spinal cord injury following aortic surgery has long been thought to be either complete motor deficit or complete absence of injury. Other organ systems susceptible to ischemia, such as the kidneys, have demonstrated that a critical mass of cellular injury is required for organ dysfunction to be apparent (Simmons et al., 2011, 2012). Previous studies of spinal cord hypoxia/ischemia have made it clear that the spinal cord neuronal injury is not an all or nothing phenomenon with injured neurons in almost all ischemic groups. However, many of these groups had relatively normal motor function despite the clear neuronal injury (Ozkisacik et al., 2011; Smith et al., 2013). The aim of this study was to not only demonstrate that spinal cord motor neuron injury can occur in the absence of gross functional deficits, but that the evaluation of this injury may be inadequate. We hypothesize that mice subjected to minimal ischemic time will have functional and histological changes despite the gross appearance of normal function.
2.
Results
2.1.
Qualitative functional assessment
2.3.
Histological
All Mice that underwent thoracic aortic occlusion had significant cyto-architectural decay with increased vacuolization, pyknotic nuclei and a significant disruption of the greywhite matter junction. Though these findings were more pronounced in paraplegic mice, they were ubiquitous in all mice following thoracic aortic occlusion (Fig. 2).
2.4.
Neuronal viability
Following sham surgery or thoracic aortic occlusion there was a step wise reduction neuronal viability. In mice subjected to IR surgery without flaccid paralysis, a significant reduction in both total anterior horn neurons (A) and lamina IX alpha-motor neurons (B) was observed. Mice with flaccid paralysis had an additional reduction in the number anterior horn motor neurons that was significant less than both other groups (Fig. 3).
3.
Discussion
The current data demonstrate several telling points with regard to the evaluation of motor function following reperfu-
Sham mice (n ¼4) had no observable functional deficits preoperatively and throughout duration of the experiment. There was no significant differences in experimental variables (weight, temperature or minimal perfusion) between mice that develop flaccid paralysis and those that were functionally normal (Table 1). Mice subjected to thoracic aortic occlusion for 3 min developed either overt paralysis (n ¼7) by 48 h or displayed minimal hind-limb deficits that were not statistically different from sham mice (n¼ 7) (Fig. 1).
2.2.
Quantitative functional assessment
While functional analysis using basso mouse score was unable to detected differences from sham mice and mice not displaying overt paralysis, single-frame motion analysis was able to detect significant differences. Assessment of hind-limb strength was performed by measuring the rump height indices of sham mice and mice that underwent IR surgery but did not exhibit gross functional deficits. Mice subjected to thoracic aortic occlusion that did not display
Fig. 1 – Functional outcomes following thoracic aortic occlusion or sham surgery. Mice that underwent thoracic aortic occlusion for three minutes had either minimal hindlimb deficits that were not statistically significant from sham mice or developed severe hind-limb deficits that were statistically significant from all other groups (*p-value o0.05).
Table 1 – Operative variables are highlighted above. Perioperative weight, in addition to average intraoperative perfusion and temperature were recorded ischemic groups. Weight was recorded preoperatively while temperature and perfusion were recorded at thirty second intervals.
Weight (gm) Average temp (1C) Minimal perfusion units (PU)
IR no paralysis
IR paralysis
p Value
27.873.1 36.47.33 57.8724.7
27.071.3 36.47.26 63.0715.8
0.54 0.97 0.64
brain research 1618 (2015) 55–60
sion of spinal cord ischemia. First, spinal cord motor neuron injury is universal with temporary aortic occlusion. Second, even the nine point Basso motor score is not sensitive enough to detect the subtle differences in neurological function resulting from ischemia and reperfusion injury. These points suggest that both diagnosis and treatment of spinal cord injury is severely lacking in patients undergoing aortic intervention. In order to further delineate spinal cord ischemia–reperfusion injury, we enlisted a murine model that reproduces the clinical entities of both immediate and delayed paraplegia seen following clinically thoracic aortic occlusion (Smith et al., 2011). This study, among others, demonstrated motor neuron degeneration in all mice undergoing ischemia and reperfusion independent of the obvious motor functional preservation (Smith et al., 2013) Thus, we began to investigate the effects of a minimal ischemic time on functional outcomes. Based on our previous studies 48 min of thoracic aortic occlusion resulted in immediate paraplegia, 4 min resulted in a delayed paraplegia and 3 min resulted in a minimal hind-limb deficit. However, the binary nature or injury, described as paraplegia or not, does not tell the whole story. When an ischemic time of 3 min was further investigated with a larger sample size we found mice exhibited a bimodal distribution of injury. After 48 h of reperfusion, mice had Table 2 – Rump height indices (RHI) were compared in sham mice and mice that did not progress to paralysis. Pre-operatively there was no difference in RHI. Following surgery, there was a significant reduction in RHI in mice that underwent thoracic aortic occlusion when compared to sham mice at all time points post-operatively. indicating a reduction in hind-limb strength. Rump Height Index
Preop 24 hrs 48 hrs 72 hrs 96 hrs
Sham
IR
P value
0.857.01 0.867.02 0.837.01 0.847.01 0.837.01
0.847.02 0.727.03 0.717.04 0.677.03 0.617.04
0.72 0.01 0.06 0.01 0.01
57
complete paralysis or minimal hind-limb deficits. Importantly, these spinal cord differences were observed in the absence of significant procedural variability. Variables including temperature, quantity of distal perfusion, and ischemic time were identical despite different functional outcomes. While the differences between the bimodal functional outcomes with similar ischemic insults deserve additional attention in future study, the current study focuses on the spinal cord changes in the functionally intact ischemic animals We next pursued the extent of histological changes in our 3 groups with a particular interest on mice with intact hindlimb function despite thoracic aortic occlusion. Mice that had been subjected to ischemia and reperfusion but did not exhibit gross functional deficits were able to ambulate despite a 40% reduction in motor neurons. Mice with a 70% reduction in motor neurons exhibited near complete function deficits. This finding suggests that once neuronal decay has extended beyond a critical mass, paraplegia will occur. Currently, only after dysfunction beyond this scope will be treated in most postoperative algorithms. In most programs, this would be limited to hypertension and CSF drainage rather than preventative measures. The current data The evolution of surgical adjuncts such as hypothermia, lumbar drains, and intercostal artery reimplantation has reduced the risk of paraplegia following aortic interventions (Okita, 2011). Unfortunately, up to 20% of patients undergoing high-risk repairs are still affected (Conrad et al., 2008; Kouchoukos et al., 2013). Most treatment algorithms for spinal cord injury are limited to extreme risk patients or patients exhibiting injury. Despite potential use of left heart bypass and hypothermia, the degree of motor neuron loss demonstrated in this study suggests the full treatment is reserved only for the worst injuries. Further treatment paradigms need to be developed for prevention in all patients at risk for several reasons. Clearly, further reduction of injury would limit exposure to a reduced life-expectancy and lifetime of debility experienced by paraplegic patients (Messé et al., 2008). The extent of injury that goes clinically undetected is unknown. More subtle injury may affect motor function over time. The relationship between functional decline and neuronal injury has been most widely studied in neurodegenerative
Fig. 2 – Representative images of anterior horns of mice in all three experimental groups (Magnification 20 ). All mice that underwent thoracic aortic occlusion had significant cyto-architectural disruption, increased vacuolization and a reduction in normal appearing anterior horn neurons. While these finding were present in all mice subjected to thoracic aortic occlusion they were much more pronounced in mice that became paraplegic. Black arrows depict injured anterior horn motor neurons.
58
brain research 1618 (2015) 55–60
Fig. 3 – Normal appearing neurons were quantified in both the anterior horn (A) and Lamina IX motor neurons (B) with representative numerical data (C). There was a significant reduction in normal appearing neurons in both groups that underwent thoracic aortic occlusion when compared to sham mice (*po0.05) Among mice that underwent thoracic aortic occlusion there was an additional reduction in anterior horn neurons in mice that became paraplegic. (þþpo0.05).
models in which clear correlation between neuronal injury and function can be defined (Golberg et al., 2011; Wang et al., 2012). Additionally, traumatic brain injury models have shown that minimal injury, not associated with functional deficits, the ischemic injury results in activation neurodegenerative cascades (Tashlykov et al., 2009). While many similarities exist between the brain and the spinal cord, the spinal cord is physiologically and functionally different than the brain. Spinal cord motor neurons represent a small population of spinal cord neurons (Bjugn, 1993) and are the largest of multipolar cells within the central nervous system (Geinismann, 1971). Additionally, motor neurons in spinal cord are much more susceptible to oxidative stress and mitochondrial dysfunction than cortical motor neurons (Panov et al., 2011). While spinal cord neuronal injury has been studied in perinatal rodents during hypoxic stress (Clancy et al., 1989), it was not been widely investigated following transient spinal cord ischemia. Only with further identification of the more subtle injury can the implications of this injury be identified. Therefore, spinal cord motor neurons may even have more detrimental long-term effects of this potentially unapparent injury. Further, we believe this model is representative of both open and endovascular treatments of complex aortic pathology. The rates of delayed paraplegia remain very similar between this approaches and both are thought to be ischemia and reperfusion in nature, with the cross clamp, clearly the ischemic induction in open cases while the creation of a watershed zone of perfusion from the excluded zone of the thoracic aorta (Eagleton and Greenberg, 2010). The perfusion to this watershed zone should improve over time, but according to Griepps group this may take up to 120 h (Etz et al., 2010). Supporting these neurons through this potential malperfusion should be inherent. However, our current treatment paradigms continue to treat the symptoms of motor neurons dysfunction rather than a protective milieu for the at risk neurons. The current study suggests that better strategies for metabolic protection of these neurons are not only necessary, but also should be applied to any patient undergoing complex aortic intervention. As treatment of interstitial and familial aortopathy presses for earlier intervention (Samadi et al., 2012), more patients will have longer post-
intervention lives with which to live with potential injury. The less pervasive deficits are not supported by current algorithm, so it could be argued that most any aortic intervention could benefit from preventative therapies. This study, however, is not without its limitations. The model has been criticized for the extent of injury relative to the short ischemic time. While the ischemic time is short relative to clinical ischemia, the metabolic rate of mice is much higher than humans (Kleiber, 1975). Further, there is no adjunctive protection from the ischemia in the model such as distal perfusion resulting in a higher ischemic burden for the model. In addition, we have followed the model out for up to 14 days without long term changes in neurological function. Despite the appearance of normal function, IR surgery induced marked cyto-architectural changes and neuronal degeneration. Furthermore high-definition gait analysis revealed significant changes in gait and activity following thoracic aortic occlusion. These data suggest that all patients undergoing surgery, even with short ischemic times, may have spinal cord injury that is not evident clinically. In summary, spinal cord injury may be ubiquitous in aortic intervention necessitating the development of neuroprotective therapies for all these procedures instead of limiting them to patients at high risk for injury or demonstrating gross deficits by exam.
4.
Experimental procedures
4.1.
Animal procedures
The Animal Care and Use Committee at the University of Colorado at Denver Health Sciences Center approved all experiments, and this investigation conformed to the Guide for the Care and Use of Laboratory Animals published by the National Institute of Health (www.nap.edu/catalog/5140.html). Adult, male C57BL/6mice were used for all experimentations.
4.2. Aortic cross clamping: Ischemia–reperfusion (IR) surgery Mice were anesthetized using 2% isoflurane and placed in the supine position. Surgery was performed under normothermic
brain research 1618 (2015) 55–60
conditions with core body temperature maintained at 36.570.5 1C using a rectal temperature probe and automatic temperature adjusting bed (Vestavia Scientific, Birmingham, AL). The aortic arch was exposed using a cervicothoracic approach as previously described. A disruption arterial flow to the spinal cord was achieved by placing vascular clamps on the aortic arch distal to the left common carotid artery and the subclavian artery for 3 min. A laser doppler blood flow monitor (Moor Instruments, UK) measuring femoral artery flow in perfusion units (PU) was placed over the left femoral artery. Prior to aortic/subclavian occlusion max perfusion of greater than 900 PU was documented. A 90% reduction in distal flow (less than 90 PU) during aortic occlusion was obtained in all mice used in ischemic groups. If mice did not have a document decreased in distal flow of more than 90% they were humanely euthanized and not included in functional outcomes.
4.3.
Qualitative functional assessment
The Basso Mouse Scale for locomotion (Basso et al., 2006) which ranges from a score of 0 for complete paraplegia to a score of 9 for normal function was used to quantify hind-limb function in mice after ischemia. Functional scores were recorded preoperatively and at 24, 48, 72 and 96 h following surgery.
4.4.
Quantitative functional assessment
Mice without overt functional deficits (Sham and IR) were place on a walking beam apparatus to analyze gait are previously described (Semler et al., 2011). Images were recorded by a high-definition camera (Panasonic HC-V500M, Newark, NJ). Single-frame images were analyzed using HD Writer AE 4.0 (Panasonic, Newark, NJ) and processed by Image Tool 3.0 (UT Health Science Center, San Antonio, TX). Rump height indices were used for calculation of hind-limb strength. This calculation was obtained by dividing the recorded height of the base of the tail above the surface of the beam by the width recorded of the beam. Three individual measurements were taking at 24 h apart by a blinded observer and averaged.
4.5.
Histological analysis
Ninety-six hrs after surgery, the animals were sacrificed. The vertebral column was removed en bloc fromT8-L3. Spinal cords were removed from the vertebral column by injection of phosphate buffered saline (PBS, pH 7.4) into the spinal column. Spinal cords were then transferred to 4% formalin where they remained for at least 24 h prior to paraffin embedding, sectioning, and hematoxylin and eosin staining. Sectioning and staining was performed by a blinded researcher who provided cranial and caudal representations of each specimen.
4.6.
Quantification of neuronal injury
Sections were evaluated by a blindly observer for neuronal viability and quantified as neurons per anterior horn. Neurons
59
within the anterior horn that contained prominent nucleoli and loose chromatin were considered normal (Kurita et al., 2005). Quantification was performed for both total anterior horn neurons and motor neurons in Lamina IX neurons.
4.7.
Statistical analysis
Statistical analysis was performed using repeated measures and the Kruskal-Wallis nonparametric ANOVA with Bonferroni/Dunn post-hoc test using, StatView (SAS Institute Inc., Cary, CA). Data are presented as mean7standard error. A pvalue o0.05 was considered significant for all statistical comparisons.
r e f e r e nc e s
Basso, D.M., Fisher, L.C., Anderson, A.J., Jakeman, L.B., McTigue, D. M., Popovich, P.G., 2006. Basso Mouse Scale for locomotion detects differences in recovery after spinal cord injury in five common mouse strains. J. Neurotrauma 23, 635–659. Becker, D.A., McGarvey, M.L., Rojvirat, C., Bavaria, J.E., Messe´, S.R., 2013. Predictors of outcome in patients with spinal cord ischemia after open aortic repair. Neurocrit. Care 18, 70–74. Bjugn, R., 1993. The use of the optical disector to estimate the number of neurons, glial and endothelial cells in the spinal cord of the mouse—with a comparative note on the rat spinal cord. Brain Res. 627, 25–33. Clancy, R.R., Sladky, J.T., Rorke, L.B., 1989. Hypoxic–ischemic spinal cord injury following perinatal asphyxia. Ann. Neurol. 25, 185–189. Conrad, M.F., Crawford, R.S., Davison, J.K., Cambria, R.P., 2008. Thoracoabdominal aneurysm repair: a 20-year perspective S856-S861. Ann. Thorac. Surg. 8. Eagleton, M.J., Greenberg, R.K., 2010. Spinal and visceral ischemia after endovascular aortic repair. J. Cardiovasc. Surg. 51, 71–83. Etz, C.D., Zoli, S., Bischoff, M.S., Bodian, C., Di Luozzo, G., Griepp, R.B., 2010. Measuring the collateral network pressure to minimize paraplegia risk in thoracoabdominal aneurysm resection. J. Thorac. Cardiovasc. Surg. 140, S125–S130. Geinismann, Y.Y., 1971. Determination of nerve and glial cell volumes in histological section. Brain Res. 26, 235–246. Golberg, N.R., Haack, A.K., Lims, N.S., Janson, O.K., Meshul, C.K., 2011. Dopaminergic and beharvioral corraletes of progressive lesioning of the nigrostriatal pathway with 1-methyl-4phenyl-1,2,3,6 tetrahydropyridine. Neuroscience 180, 256–271. Kleiber, M., 1975. The metabolic rate per unit body weight. J. Theor. Biol. 53, 199–204. Kouchoukos, N.T., Kulik, A., Castner, C.F., 2013. Outcomes after thoracoabdominal aortic aneurysm repair using hypothermic circulatory arrest. J. Thorac. Cardiovas. Surg 145, S139–S141. Kurita, N., Kawaguchi, M., Horiuchi, T., Inoue, S., Sakamoto, T., Nakamura, M., Konishi, N., Furuya, H., 2005. An evaluation of white matter injury after spinal cord ischemia in rats: a comparison with gray matter injury. Anesth. Analg. 100, 847–854. Messe´, S.R., Bavaria, J.E., Mullen, M., Cheung, A.T., Augoustides, J. G., Futsche, J., Woo, E.Y., Szeto, W.Y., Pochettino, A., Woo, Y.J., Kasner, S.E., McCarvey, M., 2008. Neurologic outcomes from high risk descending thoracic and thoracoabdominal aortic operations in the era of endovascular repair. Neurocrit. Care 9, 344–351. Okita, Y., 2011. Fighting spinal cord complication during surgery for thoracoabdominal aortic disease. Gen. Thorac. Cardiovasc. Surg. 59, 79–90.
60
brain research 1618 (2015) 55–60
Ozkisacik, E.A., Discigil, B., Boga, M., Grucun, U., Badak, M.I., Kurtoglu, T., Dikicioglu, E., 2011. Effects of cyclosporin a on neurological outcome nd serum biomarkers in the same setting of spinal cord ischemia model. Ann. Vasc. Surg. 20, 243–249. Panov, A.V., Kubalik, N., Zinchenko, N., Ridings, D.M., Radoff, D.A., Hemendinger, R., Brooks, B.R., Bonkosvsky, H.L., 2011. Metabolic and functional differences between brain and spinal cord mitochondria underlie different predisposition to pathology. Am. J. Physiol. Regul. Integr. Comp. Physiol 300, R844–R854. Samadi, A., Detaint, D., Roy, C., Arnoult, F., Delorme, G., Gautier, M., Milleron, O., Raoux, F., Meuleman, C., Hvass, U., Hamroun, D., Beroud, C., Tubach, R., Boileau, C., Jondeau, G., 2012. Surgical management of patients with Marfan syndrome: evolution throughout the years. Arch. Cardiovasc. Dis. 105, 84–90. Simmons, M.N., Fergany, A.F., Campbell, S.C., 2011. Effect of parenchymal volume preservation on kidney function after partial nephrectomy. J. Urol. 186, 405–410. Simmons, M.N., Hillyer, S.P., Lee, B.H., Fergany, A.F., Kaouk, J., Campbell, S.C., 2012. Functional recovery after partial nephrectomy: effects of volume loss and ischemic injury. J. Urol. 187, 1667–1673.
Semler, J., Wellmann, K., Wirth, F., Stein, G., Angelova, S., Ashrafi, M., Schempf, G., Ankerne, J., Ozsoy, O., Ozsoy, U., Angelov, D. N., Irintchev, A., 2011. Objective measures of motor dysfunction after compression spinal cord injury in adult rats: correlations with locomotor rating scores. J. Neurotrauma 28, 1247–1258. Smith, P.D., Puskas, F., Meng, X., Cho, D., Cleveland, J.C., Jr., Weyant, M.J., Fullerton, D.A., Reece, T.B., 2011. Ischemic dose– response in the spinal cord: both immediate and delayed paraplegia. J. Surg. Res. 7, 1–7. Smith, P.D., Bell, M.T., Puskas, F., Meng, X., Cleveland Jr., J.C., Weyant, M.J., Fullerton, D.A., Reece, T.B., 2013. Preservation of motor function after spinal cord ischemia and reperfusion injury through microglial inhibition. Ann. Thorac. Surg. 95, 1647–1653. Tashlykov, V., Katz, Y., Volkov, A., Gazit, V., Schreiber, S., Zohar, O., Pick, C.G., 2009. Minimal traumatic brain injury induced apoptotic cell death in mice. J. Mol. Neurosci. 37, 16–24. Wang, X.H., Lu, G., Hu, X., Tsang, K.S., Kwong, W.H., Wu, F.X., Meng, H.W., Jian, S., Liu, S.W., Ng, H.K., Poon, W.S., 2012. Progressive assesment of gait and neurochemial correlation in a classical murine model of Parkinson’s disease. BMC Neruosci 13, 142.
