Eur Spine J DOI 10.1007/s00586-015-3924-z
ORIGINAL ARTICLE
Neurophysiological monitoring during acute and progressive experimentally induced compression injury of the spinal cord in pigs Elena Montes1 • Jesu´s Burgos2 • Carlos Barrios3 • Gema de Blas1 • Eduardo Hevia4 • Jero´nimo Forteza5
Received: 16 December 2014 / Revised: 31 March 2015 / Accepted: 1 April 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose To evaluate the degree of acute or progressive lateral compression needed to cause neurologic injury to the spinal cord assessed by electrophysiological monitoring. Methods In five domestic pigs, the spinal cord was exposed and compressed between T8–T9 roots using a precise compression device. Two sticks placed on both sides of the spinal cord were sequentially brought together (0.5 mm every 2 min), causing progressive spinal cord compression. Acute compression was reproduced by a 2.5mm displacement of the sticks. Cord-to-cord evoked potentials were obtained with two epidural catheters. Results Increasing latency and decreasing amplitude of the evoked potentials were observed after a mean progressive displacement of the sticks of 3.2 ± 0.9 mm, disappearing after a mean displacement of 4.6 ± 1.2 mm. The potential returned after compression removal (16.8 ± 3.2 min). The potentials disappeared immediately after an acute compression of 2.5 ± 0.3 mm, without any sign of recovering after 30 min.
& Carlos Barrios [email protected] 1
Department of Clinical Neurophysiology, Hospital Ramo´n y Cajal, Madrid, Spain
2
Division of Pediatric Orthopedics, Hospital Ramo´n y Cajal, Madrid, Spain
3
Institute for Research on Musculoskeletal Disorders, Valencia Catholic University, Quevedo 2, 46001 Valencia, Spain
4
Spine Surgery Unit, Hospital La Fraternidad-Muprexpa, Madrid, Spain
5
Instituto Valenciano de Patologı´a, Valencia Catholic University, Valencia, Spain
Conclusions The experimental model replicates the mechanism of a spinal cord injury caused by medially displaced screws into the spinal canal. The spinal cord had more ability for adaptation to progressive and slow compression than to acute mechanisms. Keywords Spinal cord Compression injury Epidural neuromonitoring
Introduction The main objective of neurophysiological monitoring during spinal surgery is to prevent the occurrence of spinal cord injury. The incidence of neurological complications during spinal surgery is about 1 %, and a third of these cases correspond to spinal cord injury (SCI) [1]. Severe SCI may lead to a devastating loss of neurological function below the level of the injury. During surgical treatment of idiopathic scoliosis through a posterior approach, the incidence of neurological complications is relatively low accounting for 1.7 %, with 34 % being are injuries of the spinal cord [2]. The risk of spinal cord injury increases with the complexity of the surgery, which involves several predisposing factors such as spinal cord ischemia, compression during instrumentation, or the mobilization of unstable segments during surgery [3, 4]. The introduction of thoracic pedicle screws, the development of three-column osteotomies, and vertebral resection techniques have increased the number of surgeries performed using a posterior approach [5, 6]. Using these techniques, neural elements can be brought into the surgical field, such that their displacements are accompanied by an increased risk of neurological complications. During surgical maneuvers, neural damage might be produced by
123
Eur Spine J
local trauma or compression by the instrumental elements used for spine fusion purposes (screws, rods, etc.). Screw placement is particularly difficult when rotational scoliotic deformities are present [7, 8]. In fact, malpositioning is the most commonly reported complication of thoracic pedicle screw placement according to a recent systematic review on the complications of pedicle screw fixation during scoliosis surgery [9]. Screw misplacement with pedicle medial cortex violation and invasion of the spinal canal entails a high risk of potential injuries to the spinal cord and nerve roots. In the spinal cord, fibers within the pyramidal or lateral corticospinal tract are hence particularly vulnerable in medially malpositioned thoracic pedicle screws. The proximity of the pyramidal tract to the medial wall of the pedicle makes it highly susceptible to mechanical lesions [10, 11]. The involvement of this neural pathway in injury is also associated with significant residual motor deficits [11, 12]. This experimental study was aimed at establishing the degree of lateral compression tolerated by the spinal cord before causing a neurological injury. Neurophysiological monitoring was used to assess the functional neural damage. Moreover, the limits of compression tolerance were analyzed with respect to the type of the applied compression: slow or gradual, and rapid or acute. This study attempts to simulate the lateral compression mechanism that can be induced by thoracic pedicle screw misplacement during spinal surgery (Fig. 1). The experimental design included also an attempt to correlate the neuromonitoring findings during thoracic cord lateral compression with the gross histologic changes in the injured spinal cord. In addition, a new device for spinal cord lateral clip compression for experimental use with big animals such as pigs is presented.
Materials and methods The Research Ethics Committee of the University Hospital approved the study protocol. This experimental procedure is in line with current national legislation on the protection of animals used for experimental and other scientific purposes. The experimental animals under study were five domestic pigs between 29 and 39 kg in weight (35 kg average weight). The use of animals is justified for reproducing the different mechanisms or factors that can influence the spinal cord injury during spine surgery in a controlled environment and in an experimental model that is highly similar to humans. Anesthetic technique The procedure was performed under conventional general anesthesia and endotracheal intubation. The experimental animals were sedated by intramuscular injection of 10 mg/ kg ketamine and 20 mg/kg of sodium thiopental. A peripheral vein was channeled into an ear vein. 0.5 mg of SC IV atropine and 30 mg of IV ketorolac tromethamine were administered as premedication. After sedation, animals were intubated while maintaining general anesthesia with 2–2.5 % sevoflurane inhalation in 100 % of O2. Vecuronium was used as a muscle relaxant. The analgesia infusion was controlled by at 80 ml/h of a IV mixture comprising 5 mg of remifentanil hydrochloride and 0.45 mg of fentanyl diluted in 500 ml of 0.9 % saline. Mechanical ventilation was performed with continuous monitoring of O2 saturation, blood pressure and heart rate. After surgery, the animals were killed by intravenous injection of an overdose of sodium thiopental, and then they were incinerated. Surgical technique
Fig. 1 Severe invasion of the spinal canal following pedicle screw fixation producing a lateral compression of the spinal cord
123
The domestic pig was placed in a prone position. A posterior longitudinal midline incision was performed to expose the spine at thoracic level from the transverse processes of the T7 to the T11 vertebrae. A wide laminectomy was performed at those levels, exposing the spinal cord and the T8 and T9 nerve roots bilaterally. The width of the medullary canal and the dural sac was measured at the level where the compression was planned (between roots T8 and T9). To apply lateral cord compression, a pair of sticks of the compression device, which were specially designed for this study (Fig. 2), was applied on both sides of the exposed spinal cord. The degree of spinal cord compression was quantified in mm of displacement of the sticks that were brought together at a rate of 0.25 mm every 2 min to effect a progressive compression. The acute compression of the spinal cord was also reproduced by a 2.5 mm rapid displacement of the sticks (Fig. 3).
Eur Spine J Fig. 2 Details of the device used for spinal cord compression and panel for displacement control
Fig. 3 Placement of the sticks for spinal cord compression (left) and an image of the final compression (right)
Neurophysiological monitoring A keypoint equipment four-channel amplifier (Medtronic, Skovlunde, Denmark) was used for neurophysiological monitoring. Cord-to-cord motor evoked potentials were obtained with two epidural electrodes,
stimulating proximal to T6 and recording below the compression level distal to T10 for each sequential approach of the sticks. Before starting cord compression maneuvers, a control potential track was recorded as a reference to determine subsequent changes following cord compression.
123
Eur Spine J
Histology Once the experiment was concluded, the animals were killed by an intravenous injection of an overdose of sodium thiopental. A 3-cm segment of the spinal cord from three of the five animals centered on the injury site was harvested for histological analysis. These tissues were fixed for 24 h by adding 4 % formalin and then dehydrated through alcohol and xilol, and then embedded in paraffin. Finally, 4-lm cross sections were obtained and tissues were stained with hematoxylin-eosin (HE). For electron microscopy studies, the samples were post-fixed with 2 % osmium, rinsed, dehydrated and embedded in Durcupan resin (Fluka, Sigma-Aldrich, St. Louis, USA). Ultrathin sections (60–90 nanometer) were prepared with the Ultracut and stained with lead citrate. Finally, photomicrographs were obtained under a transmission electron microscope FEI Tecnai G2 Spirit (FEI Europe, Eindhoven, The Netherlands) using a digital camera Morada (Olympus Soft Image Solutions GmbH, Mu¨nster, Germany). Data management The Statistical Package for the Social Sciences version 21 (IBM, Chicago, IL, USA) was used for data processing and analysis. Quantitative variables are expressed as the mean with the standard deviation (SD) in parentheses. Using a pairwise analysis with the Wilcoxon signed ranks test, the changes over time in epidural neurophysiologic signals after spinal cord compression were studied. A p value \0.05 was considered to be statistically significant.
Results The mean width of the dural sac was 7.1 mm. During slow, progressive spinal cord compression, slight changes in the neurophysiological signals occurred after a mean compression of 3.2 mm (SD = 0.9), which means a 45 % reduction in spinal cord lateral diameter. The electrophysiological signals finally disappeared after a 4.6 mm mean displacement of the sticks of the device (SD = 1.2), implying a 64.8 % reduction of the spinal cord diameter (Fig. 4). This change was statistically significant (Wilcoxon test, Z = -2.032, p = 0.042). After the release of compression, potentials recovered with an average latency of 16.8 min (SD = 3.2). The experiment conducted by an acute compression exerted by the sticks device with an average displacement of 2.5 mm (SD = 0.3) resulted in the immediate and irreversible loss of spinal evoked potential in the five cases studied. The percentage of reduction of the latero-lateral diameter of the spinal cord was 35.2 % (Fig. 4). The
123
statistical analysis comparing the compression needed for absence of signals after acute and progressive compression disclosed also significant differences (Wilcoxon test, Z = -2.023, p = 0.043). After 30 min of electrophysiological monitoring through direct epidural stimulation and recording, there was no evidence of recovery in any of the five cases studied. Figure 5 shows representative sections of the spinal cord at the site of progressive lateral compression. There were areas with partial vacuolization of the neural tissue and pycnotic red neurons indicating a moderate ischemic injury. There was gradual increased vacuolization and axonal damage with increasing clip force. The persisting tissue showed some areas with infiltration of macrophages. The vascular structure was respected. More severe and extensive ischemic lesions were found at the site of spinal cord acute compression (Fig. 6). The vacuolization areas and the loss of structure of the myelin sheaths detected in the electron microscopy were remarkable. In the case with acute compression, a more hypoxic injury, neurons showed intense pycnotic nuclei, cytoplasmatic chromatolysis and cell necrosis.
Discussion So far, SCI has been extensively investigated in experimental models. Dynamic or static spinal cord traumatic compression has been experimentally performed through different mechanisms: inflatable balls, clamping, and dropping a weight, etc. [13]. The most common mechanism of SCI in humans is a combination of acute impact followed by persisting compression [14]; therefore, most authors has concentrated their research on the acute clip compression model in rats [15, 16]. The different devices in use for inducing SCI simulate acute antero-posterior compression but do not reproduce the effect of the persisting and progressive compression of the cord that can occur after thoracic pedicle screw misplacement [17–19]. To our knowledge, the lateral clip compression device used in this research is the first that replicates both the acute impact and the progressive lateral extradural compression of the spinal cord characteristic of the vast majority of neurological complications during surgical correction of spinal deformities. This model was especially design for experimental application on big animal such as the pig. When the clip closes rapidly, a compression force applied to both lateral aspects of the spinal cord produces an acute cord contusion. The force can be varied based on the speed and amount of closure of the blades. The device permitted a progressive compression by controlling the speed of closure.
Eur Spine J Fig. 4 Response of the electrophysiological signals to the amount of cord compression (left) and reduction of the laterlateral spinal cord in the two compression techniques (right). *Comparison between initial alterations and absence of signals in progressive compression, p \ 0.05; **comparison of the absence of signals between acute and progressive compression, p \ 0.05
Fig. 5 Spinal cord of an animal with lateral progressive compression. a Spinal cord gray matter at the level of the ventral horn showing neuronal pycnosis and cytoplasmic acidophilia with loss of Nissl granules (HE, 940). b Vacuolization of neural tissues (HE, 940)
Although many experimental models of SCI have been used, the literature is limited as to studies correlating the severity of thoracic cord compression and quantitative histologic changes in the injured spinal cord. Using rats as
Fig. 6 Spinal cord corresponding to a case of lateral acute compression. a Ultrastructural image showing nuclear pycnosis, cytoplasmic chromatolysis and cell necrosis at neuronal level. b Myelin degeneration with loss of structure and transformation into ovoid bodies
experimental animals in a dorso-ventral compression model, the severity of a thoracic SCI correlated with the histologic changes in the spinal cord and the functional scores 4 weeks after injury [19]. To our knowledge, the immediate histologic changes in the spinal cord following a lateral compression have been
123
Eur Spine J
never assessed. Most of the studies in this field, evaluate the experimental animals with different neurologic test after a several weeks after spine injury [15] Independently of the injury mechanism used, the morphological changes are quite similar [15, 19, 20]. These changes include edema of the neural tissues, being maximal at 8 h after lesion, and disruption of axons starting early after the injury [17]. Other common findings are the glial response with swollen and enlarged astrocytes, seen several weeks post lesion, and the final cavitation of the spinal cord in relation to the amount of necrosis induced by the injury [19, 20]. Cavity volume is commonly used in SCI research as a morphologic outcome measure and a tool for assessing severity of injury because it correlates positively with locomotor function [18–22]. The present study describes for the first time both neurophysiologic and histologic outcomes following thoracic SCI induced by either an acute or progressive extradural lateral compression model in the pig. In the first stage of this study, the slow progressive compression was removed immediately after recording a loss of electrophysiological signals. In all of the cases, the potentials recovered completely in a few minutes. These findings agree with previous studies showing that the ability for electrophysiological recovery is related to the duration of the compression to the spinal cord [15]. This feature suggests that the spinal cord has a high tolerance to slow and progressive compression. However, when rapid and acute compression was applied and released immediately after observing changes in the electrophysiological signals, there was no recovery in any of the experimental animals. This fact suggests that the spinal cord does not tolerate the acute mechanism of injury. During human spine surgeries performed by a posterior approach, compression tends to result from poor positioning of the pedicle screws violating the spinal canal. In this experimental model, reproducing this mechanism of lateral spinal cord compression, the evoked potential was directly elicited at the spinal cord by placing one electrode at the proximal region and another in the distal spinal cord. Regarding histological findings, this experiment was acute and the spinal cord samples were collected immediately after the procedure was completed. There was no enough time to detect necrotic areas and cavitation. However, an interesting aspect of this study lies on the detection of existing morphological changes at the time of neurophysiological signal impairment. That changes can be considered as the early stages of an ischemic SCI together with the mechanical compromise of neural tissue. The early structural axonal degeneration may be linked to the decrease and the loss of electrophysiological potentials. Several limitations in this work should be noted. The first is that this study was conducted in experimental animals (young domestic pigs), so caution should be maximized when
123
attempting to transfer the results to clinical practice, since this work obviously cannot be replicated in humans. It seems reasonable to assume that the tolerance of the spinal cord compression mechanism in these healthy young animals may be greater than in other older animals or those with a vascular pathology. Another limitation is related to the method employed for spinal cord compression. Compression was applied in a slow and controlled manner to establish the same type of measurement and ensure that the study was reproducible. However, this is not the case in clinical practice. Finally, this study was conducted on experimental animals with a healthy spine without scoliotic deformities or severe kyphosis. In these conditions, the spinal cord is more sensitive to the different mechanisms of injury in areas where the scoliotic curvature is greater, especially during correction maneuvers. It is possibly necessary for this study to be reproducible in experimental animals with spinal deformities to confirm these findings in situations as similar as possible to the usual practice. In summary, using an experimental pig model, the results show that the spinal cord is highly sensitive to acute compression, causing an irreversible loss and complete cord-to-cord evoked potentials. However, the spinal cord seems to have more tolerance to a slow, gradual compression and the electrophysiological responses can be absent just after the initial injury. From a clinical point of view, the results confirm the need to avoid sudden maneuvers of cord mobilization or causing acute trauma to the spinal cord, even with low intensity. Finally, the experimental model will be useful for planning future studies. Conflict of interest of interest.
Authors state that there is no potential conflict
References 1. Hamilton DK, Smith JS, Sansur CA, Glassman SD, Ames CP, Berven SH, Polly DW Jr, Perra JH, Knapp DR, Boachie-Adjei O, McCarthy RE, Shaffrey CI (2011) Rates of new neurological deficit associated with spine surgery based on 108,419 procedures: a report of the Scoliosis Research Society Morbidity and Mortality Committee. Spine 36:1218–1228 2. Coe JD, Arlet V, Donaldson W, Berven S, Hanson DS, Mudiyam R, Perra JH, Shaffrey CI (2006) Complications in spinal fusion for adolescent idiopathic scoliosis in the new millennium. A report of the Scoliosis Research Society Morbidity and Mortality Committee. Spine 31:345–349 3. Cheh G, Lenke LG, Padberg AM, Kim YJ, Daubs MD, Kuhns C, Stobbs G, Hensley M (2008) Loss of spinal cord monitoring signals in children during thoracic kyphosis correction with spinal osteotomy: why does it occur and what should you do? Spine 33:1093–1099 4. Suk SI, Chung ER, Kim JH, Kim SS, Lee JS, Choi WK (2005) Posterior vertebral column resection for severe rigid scoliosis. Spine 30:1682–1687 5. Lenke LG, O’Leary PT, Bridwell KH, Sides BA, Koester LA, Blanke KM (2009) Posterior vertebral column resection for
Eur Spine J
6.
7.
8.
9.
10.
11.
12.
13.
severe pediatric deformity: minimum two-year follow-up of thirty-five consecutive patients. Spine 34:2213–2221 Hsieh PC, Li KW, Sciubba DM, Suk I, Wolinsky JP, Gokaslan ZL (2009) Posterior-only approach for total en bloc spondylectomy for malignant primary spinal neoplasms: anatomic considerations and operative nuances. Neurosurgery 65(6 Suppl):173–181 Liljenqvist UR, Halm HF, Link TM (1997) Pedicle screw instrumentation of the thoracic spine in idiopathic scoliosis. Spine 22:2239–2245 Suk SI, Lee CK, Kim WJ et al (1995) Segmental pedicle screw fixation in the treatment of thoracic idiopathic scoliosis. Spine 20:1399–1405 Hicks JM, Singla A, Shen FH et al (2010) Complications of pedicle screw fixation in scoliosis surgery: a systematic review. Spine 35:465–470 Blight AR, Decrescito V (1986) Morphometric analysis of experimental spinal cord injury in the cat: the relation of injury intensity to survival of myelinated axons. Neuroscience 19:321–341 Quencer RM, Bunge RP, Egnor M et al (1992) Acute traumatic central cord syndrome: MRI-pathological correlations. Neuroradiology 34:85–94 Donohue ML, Murtagh-Schaffer C, Basta J et al (2008) Pulsetrain stimulation for detecting medial malpositioning of thoracic pedicle screws. Spine 33:378–385 Falconer JC, Narayana PA, Bhattacharjee M, Liu SJ (1996) Characterization of an experimental spinal cord injury model using waveform and morphometric analysis. Spine 21:104–112
14. De Girolami U, Frosch MP, Tator CH (2002) Regional neuropathology diseases of the spinal cord and vertebral column. In: Graham DI, Lantos PL (eds) Greenfield’s neuropathology, 7th edn. Arnold, London, pp 1063–1101 15. Rivlin AS, Tator CH (1978) Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surg Neurol 10:38–43 16. Fehlings MG, Tator CH, Linden RD (1989) The relationships among the severity of spinal cord injury, motor and somatosensory evoked potentials and spinal cord blood flow. Electroencephalogr Clin Neurophysiol 74:241–259 17. Cao Q, Zhang YP, Iannotti C et al (2005) Functional and electrophysiological changes after graded traumatic spinal cord injury in adult rat. Exp Neurol 191(suppl 1):3–16 18. Scheff SW, Rabchevsky AG, Fugaccia I et al (2003) Experimental modeling of spinal cord injury: characterization of a force-defined injury device. J Neurotrauma 20:179–193 19. Poon PC, Gupta D, Shoichet MS, Tator CH (2007) Clip compression model is useful for thoracic spinal cord injuries. Histologic and functional correlates. Spine 32:2853–2855 20. Noble LJ, Wrathall JR (1989) Correlative analysis of lesion development and functional status after graded spinal cord contusive injuries in the rat. Exp Neurol 103:34–40 21. Carlson GD, Gorden CD, Oliff HS, Pillai JJ, LaManna JC (2003) Sustained spinal cord compression: part I: time-dependent effect on long-term pathophysiology. J Bone Joint Surg Am 85-A:86–94 ˚ , Sundstro¨m E (1997) Clip compression 22. von Euler M, Seiger A injury in the spinal cord: a correlative study of neurological and morphological alterations. Exp Neurol 145:502–510
123
ORIGINAL ARTICLE
Neurophysiological monitoring during acute and progressive experimentally induced compression injury of the spinal cord in pigs Elena Montes1 • Jesu´s Burgos2 • Carlos Barrios3 • Gema de Blas1 • Eduardo Hevia4 • Jero´nimo Forteza5
Received: 16 December 2014 / Revised: 31 March 2015 / Accepted: 1 April 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose To evaluate the degree of acute or progressive lateral compression needed to cause neurologic injury to the spinal cord assessed by electrophysiological monitoring. Methods In five domestic pigs, the spinal cord was exposed and compressed between T8–T9 roots using a precise compression device. Two sticks placed on both sides of the spinal cord were sequentially brought together (0.5 mm every 2 min), causing progressive spinal cord compression. Acute compression was reproduced by a 2.5mm displacement of the sticks. Cord-to-cord evoked potentials were obtained with two epidural catheters. Results Increasing latency and decreasing amplitude of the evoked potentials were observed after a mean progressive displacement of the sticks of 3.2 ± 0.9 mm, disappearing after a mean displacement of 4.6 ± 1.2 mm. The potential returned after compression removal (16.8 ± 3.2 min). The potentials disappeared immediately after an acute compression of 2.5 ± 0.3 mm, without any sign of recovering after 30 min.
& Carlos Barrios [email protected] 1
Department of Clinical Neurophysiology, Hospital Ramo´n y Cajal, Madrid, Spain
2
Division of Pediatric Orthopedics, Hospital Ramo´n y Cajal, Madrid, Spain
3
Institute for Research on Musculoskeletal Disorders, Valencia Catholic University, Quevedo 2, 46001 Valencia, Spain
4
Spine Surgery Unit, Hospital La Fraternidad-Muprexpa, Madrid, Spain
5
Instituto Valenciano de Patologı´a, Valencia Catholic University, Valencia, Spain
Conclusions The experimental model replicates the mechanism of a spinal cord injury caused by medially displaced screws into the spinal canal. The spinal cord had more ability for adaptation to progressive and slow compression than to acute mechanisms. Keywords Spinal cord Compression injury Epidural neuromonitoring
Introduction The main objective of neurophysiological monitoring during spinal surgery is to prevent the occurrence of spinal cord injury. The incidence of neurological complications during spinal surgery is about 1 %, and a third of these cases correspond to spinal cord injury (SCI) [1]. Severe SCI may lead to a devastating loss of neurological function below the level of the injury. During surgical treatment of idiopathic scoliosis through a posterior approach, the incidence of neurological complications is relatively low accounting for 1.7 %, with 34 % being are injuries of the spinal cord [2]. The risk of spinal cord injury increases with the complexity of the surgery, which involves several predisposing factors such as spinal cord ischemia, compression during instrumentation, or the mobilization of unstable segments during surgery [3, 4]. The introduction of thoracic pedicle screws, the development of three-column osteotomies, and vertebral resection techniques have increased the number of surgeries performed using a posterior approach [5, 6]. Using these techniques, neural elements can be brought into the surgical field, such that their displacements are accompanied by an increased risk of neurological complications. During surgical maneuvers, neural damage might be produced by
123
Eur Spine J
local trauma or compression by the instrumental elements used for spine fusion purposes (screws, rods, etc.). Screw placement is particularly difficult when rotational scoliotic deformities are present [7, 8]. In fact, malpositioning is the most commonly reported complication of thoracic pedicle screw placement according to a recent systematic review on the complications of pedicle screw fixation during scoliosis surgery [9]. Screw misplacement with pedicle medial cortex violation and invasion of the spinal canal entails a high risk of potential injuries to the spinal cord and nerve roots. In the spinal cord, fibers within the pyramidal or lateral corticospinal tract are hence particularly vulnerable in medially malpositioned thoracic pedicle screws. The proximity of the pyramidal tract to the medial wall of the pedicle makes it highly susceptible to mechanical lesions [10, 11]. The involvement of this neural pathway in injury is also associated with significant residual motor deficits [11, 12]. This experimental study was aimed at establishing the degree of lateral compression tolerated by the spinal cord before causing a neurological injury. Neurophysiological monitoring was used to assess the functional neural damage. Moreover, the limits of compression tolerance were analyzed with respect to the type of the applied compression: slow or gradual, and rapid or acute. This study attempts to simulate the lateral compression mechanism that can be induced by thoracic pedicle screw misplacement during spinal surgery (Fig. 1). The experimental design included also an attempt to correlate the neuromonitoring findings during thoracic cord lateral compression with the gross histologic changes in the injured spinal cord. In addition, a new device for spinal cord lateral clip compression for experimental use with big animals such as pigs is presented.
Materials and methods The Research Ethics Committee of the University Hospital approved the study protocol. This experimental procedure is in line with current national legislation on the protection of animals used for experimental and other scientific purposes. The experimental animals under study were five domestic pigs between 29 and 39 kg in weight (35 kg average weight). The use of animals is justified for reproducing the different mechanisms or factors that can influence the spinal cord injury during spine surgery in a controlled environment and in an experimental model that is highly similar to humans. Anesthetic technique The procedure was performed under conventional general anesthesia and endotracheal intubation. The experimental animals were sedated by intramuscular injection of 10 mg/ kg ketamine and 20 mg/kg of sodium thiopental. A peripheral vein was channeled into an ear vein. 0.5 mg of SC IV atropine and 30 mg of IV ketorolac tromethamine were administered as premedication. After sedation, animals were intubated while maintaining general anesthesia with 2–2.5 % sevoflurane inhalation in 100 % of O2. Vecuronium was used as a muscle relaxant. The analgesia infusion was controlled by at 80 ml/h of a IV mixture comprising 5 mg of remifentanil hydrochloride and 0.45 mg of fentanyl diluted in 500 ml of 0.9 % saline. Mechanical ventilation was performed with continuous monitoring of O2 saturation, blood pressure and heart rate. After surgery, the animals were killed by intravenous injection of an overdose of sodium thiopental, and then they were incinerated. Surgical technique
Fig. 1 Severe invasion of the spinal canal following pedicle screw fixation producing a lateral compression of the spinal cord
123
The domestic pig was placed in a prone position. A posterior longitudinal midline incision was performed to expose the spine at thoracic level from the transverse processes of the T7 to the T11 vertebrae. A wide laminectomy was performed at those levels, exposing the spinal cord and the T8 and T9 nerve roots bilaterally. The width of the medullary canal and the dural sac was measured at the level where the compression was planned (between roots T8 and T9). To apply lateral cord compression, a pair of sticks of the compression device, which were specially designed for this study (Fig. 2), was applied on both sides of the exposed spinal cord. The degree of spinal cord compression was quantified in mm of displacement of the sticks that were brought together at a rate of 0.25 mm every 2 min to effect a progressive compression. The acute compression of the spinal cord was also reproduced by a 2.5 mm rapid displacement of the sticks (Fig. 3).
Eur Spine J Fig. 2 Details of the device used for spinal cord compression and panel for displacement control
Fig. 3 Placement of the sticks for spinal cord compression (left) and an image of the final compression (right)
Neurophysiological monitoring A keypoint equipment four-channel amplifier (Medtronic, Skovlunde, Denmark) was used for neurophysiological monitoring. Cord-to-cord motor evoked potentials were obtained with two epidural electrodes,
stimulating proximal to T6 and recording below the compression level distal to T10 for each sequential approach of the sticks. Before starting cord compression maneuvers, a control potential track was recorded as a reference to determine subsequent changes following cord compression.
123
Eur Spine J
Histology Once the experiment was concluded, the animals were killed by an intravenous injection of an overdose of sodium thiopental. A 3-cm segment of the spinal cord from three of the five animals centered on the injury site was harvested for histological analysis. These tissues were fixed for 24 h by adding 4 % formalin and then dehydrated through alcohol and xilol, and then embedded in paraffin. Finally, 4-lm cross sections were obtained and tissues were stained with hematoxylin-eosin (HE). For electron microscopy studies, the samples were post-fixed with 2 % osmium, rinsed, dehydrated and embedded in Durcupan resin (Fluka, Sigma-Aldrich, St. Louis, USA). Ultrathin sections (60–90 nanometer) were prepared with the Ultracut and stained with lead citrate. Finally, photomicrographs were obtained under a transmission electron microscope FEI Tecnai G2 Spirit (FEI Europe, Eindhoven, The Netherlands) using a digital camera Morada (Olympus Soft Image Solutions GmbH, Mu¨nster, Germany). Data management The Statistical Package for the Social Sciences version 21 (IBM, Chicago, IL, USA) was used for data processing and analysis. Quantitative variables are expressed as the mean with the standard deviation (SD) in parentheses. Using a pairwise analysis with the Wilcoxon signed ranks test, the changes over time in epidural neurophysiologic signals after spinal cord compression were studied. A p value \0.05 was considered to be statistically significant.
Results The mean width of the dural sac was 7.1 mm. During slow, progressive spinal cord compression, slight changes in the neurophysiological signals occurred after a mean compression of 3.2 mm (SD = 0.9), which means a 45 % reduction in spinal cord lateral diameter. The electrophysiological signals finally disappeared after a 4.6 mm mean displacement of the sticks of the device (SD = 1.2), implying a 64.8 % reduction of the spinal cord diameter (Fig. 4). This change was statistically significant (Wilcoxon test, Z = -2.032, p = 0.042). After the release of compression, potentials recovered with an average latency of 16.8 min (SD = 3.2). The experiment conducted by an acute compression exerted by the sticks device with an average displacement of 2.5 mm (SD = 0.3) resulted in the immediate and irreversible loss of spinal evoked potential in the five cases studied. The percentage of reduction of the latero-lateral diameter of the spinal cord was 35.2 % (Fig. 4). The
123
statistical analysis comparing the compression needed for absence of signals after acute and progressive compression disclosed also significant differences (Wilcoxon test, Z = -2.023, p = 0.043). After 30 min of electrophysiological monitoring through direct epidural stimulation and recording, there was no evidence of recovery in any of the five cases studied. Figure 5 shows representative sections of the spinal cord at the site of progressive lateral compression. There were areas with partial vacuolization of the neural tissue and pycnotic red neurons indicating a moderate ischemic injury. There was gradual increased vacuolization and axonal damage with increasing clip force. The persisting tissue showed some areas with infiltration of macrophages. The vascular structure was respected. More severe and extensive ischemic lesions were found at the site of spinal cord acute compression (Fig. 6). The vacuolization areas and the loss of structure of the myelin sheaths detected in the electron microscopy were remarkable. In the case with acute compression, a more hypoxic injury, neurons showed intense pycnotic nuclei, cytoplasmatic chromatolysis and cell necrosis.
Discussion So far, SCI has been extensively investigated in experimental models. Dynamic or static spinal cord traumatic compression has been experimentally performed through different mechanisms: inflatable balls, clamping, and dropping a weight, etc. [13]. The most common mechanism of SCI in humans is a combination of acute impact followed by persisting compression [14]; therefore, most authors has concentrated their research on the acute clip compression model in rats [15, 16]. The different devices in use for inducing SCI simulate acute antero-posterior compression but do not reproduce the effect of the persisting and progressive compression of the cord that can occur after thoracic pedicle screw misplacement [17–19]. To our knowledge, the lateral clip compression device used in this research is the first that replicates both the acute impact and the progressive lateral extradural compression of the spinal cord characteristic of the vast majority of neurological complications during surgical correction of spinal deformities. This model was especially design for experimental application on big animal such as the pig. When the clip closes rapidly, a compression force applied to both lateral aspects of the spinal cord produces an acute cord contusion. The force can be varied based on the speed and amount of closure of the blades. The device permitted a progressive compression by controlling the speed of closure.
Eur Spine J Fig. 4 Response of the electrophysiological signals to the amount of cord compression (left) and reduction of the laterlateral spinal cord in the two compression techniques (right). *Comparison between initial alterations and absence of signals in progressive compression, p \ 0.05; **comparison of the absence of signals between acute and progressive compression, p \ 0.05
Fig. 5 Spinal cord of an animal with lateral progressive compression. a Spinal cord gray matter at the level of the ventral horn showing neuronal pycnosis and cytoplasmic acidophilia with loss of Nissl granules (HE, 940). b Vacuolization of neural tissues (HE, 940)
Although many experimental models of SCI have been used, the literature is limited as to studies correlating the severity of thoracic cord compression and quantitative histologic changes in the injured spinal cord. Using rats as
Fig. 6 Spinal cord corresponding to a case of lateral acute compression. a Ultrastructural image showing nuclear pycnosis, cytoplasmic chromatolysis and cell necrosis at neuronal level. b Myelin degeneration with loss of structure and transformation into ovoid bodies
experimental animals in a dorso-ventral compression model, the severity of a thoracic SCI correlated with the histologic changes in the spinal cord and the functional scores 4 weeks after injury [19]. To our knowledge, the immediate histologic changes in the spinal cord following a lateral compression have been
123
Eur Spine J
never assessed. Most of the studies in this field, evaluate the experimental animals with different neurologic test after a several weeks after spine injury [15] Independently of the injury mechanism used, the morphological changes are quite similar [15, 19, 20]. These changes include edema of the neural tissues, being maximal at 8 h after lesion, and disruption of axons starting early after the injury [17]. Other common findings are the glial response with swollen and enlarged astrocytes, seen several weeks post lesion, and the final cavitation of the spinal cord in relation to the amount of necrosis induced by the injury [19, 20]. Cavity volume is commonly used in SCI research as a morphologic outcome measure and a tool for assessing severity of injury because it correlates positively with locomotor function [18–22]. The present study describes for the first time both neurophysiologic and histologic outcomes following thoracic SCI induced by either an acute or progressive extradural lateral compression model in the pig. In the first stage of this study, the slow progressive compression was removed immediately after recording a loss of electrophysiological signals. In all of the cases, the potentials recovered completely in a few minutes. These findings agree with previous studies showing that the ability for electrophysiological recovery is related to the duration of the compression to the spinal cord [15]. This feature suggests that the spinal cord has a high tolerance to slow and progressive compression. However, when rapid and acute compression was applied and released immediately after observing changes in the electrophysiological signals, there was no recovery in any of the experimental animals. This fact suggests that the spinal cord does not tolerate the acute mechanism of injury. During human spine surgeries performed by a posterior approach, compression tends to result from poor positioning of the pedicle screws violating the spinal canal. In this experimental model, reproducing this mechanism of lateral spinal cord compression, the evoked potential was directly elicited at the spinal cord by placing one electrode at the proximal region and another in the distal spinal cord. Regarding histological findings, this experiment was acute and the spinal cord samples were collected immediately after the procedure was completed. There was no enough time to detect necrotic areas and cavitation. However, an interesting aspect of this study lies on the detection of existing morphological changes at the time of neurophysiological signal impairment. That changes can be considered as the early stages of an ischemic SCI together with the mechanical compromise of neural tissue. The early structural axonal degeneration may be linked to the decrease and the loss of electrophysiological potentials. Several limitations in this work should be noted. The first is that this study was conducted in experimental animals (young domestic pigs), so caution should be maximized when
123
attempting to transfer the results to clinical practice, since this work obviously cannot be replicated in humans. It seems reasonable to assume that the tolerance of the spinal cord compression mechanism in these healthy young animals may be greater than in other older animals or those with a vascular pathology. Another limitation is related to the method employed for spinal cord compression. Compression was applied in a slow and controlled manner to establish the same type of measurement and ensure that the study was reproducible. However, this is not the case in clinical practice. Finally, this study was conducted on experimental animals with a healthy spine without scoliotic deformities or severe kyphosis. In these conditions, the spinal cord is more sensitive to the different mechanisms of injury in areas where the scoliotic curvature is greater, especially during correction maneuvers. It is possibly necessary for this study to be reproducible in experimental animals with spinal deformities to confirm these findings in situations as similar as possible to the usual practice. In summary, using an experimental pig model, the results show that the spinal cord is highly sensitive to acute compression, causing an irreversible loss and complete cord-to-cord evoked potentials. However, the spinal cord seems to have more tolerance to a slow, gradual compression and the electrophysiological responses can be absent just after the initial injury. From a clinical point of view, the results confirm the need to avoid sudden maneuvers of cord mobilization or causing acute trauma to the spinal cord, even with low intensity. Finally, the experimental model will be useful for planning future studies. Conflict of interest of interest.
Authors state that there is no potential conflict
References 1. Hamilton DK, Smith JS, Sansur CA, Glassman SD, Ames CP, Berven SH, Polly DW Jr, Perra JH, Knapp DR, Boachie-Adjei O, McCarthy RE, Shaffrey CI (2011) Rates of new neurological deficit associated with spine surgery based on 108,419 procedures: a report of the Scoliosis Research Society Morbidity and Mortality Committee. Spine 36:1218–1228 2. Coe JD, Arlet V, Donaldson W, Berven S, Hanson DS, Mudiyam R, Perra JH, Shaffrey CI (2006) Complications in spinal fusion for adolescent idiopathic scoliosis in the new millennium. A report of the Scoliosis Research Society Morbidity and Mortality Committee. Spine 31:345–349 3. Cheh G, Lenke LG, Padberg AM, Kim YJ, Daubs MD, Kuhns C, Stobbs G, Hensley M (2008) Loss of spinal cord monitoring signals in children during thoracic kyphosis correction with spinal osteotomy: why does it occur and what should you do? Spine 33:1093–1099 4. Suk SI, Chung ER, Kim JH, Kim SS, Lee JS, Choi WK (2005) Posterior vertebral column resection for severe rigid scoliosis. Spine 30:1682–1687 5. Lenke LG, O’Leary PT, Bridwell KH, Sides BA, Koester LA, Blanke KM (2009) Posterior vertebral column resection for
Eur Spine J
6.
7.
8.
9.
10.
11.
12.
13.
severe pediatric deformity: minimum two-year follow-up of thirty-five consecutive patients. Spine 34:2213–2221 Hsieh PC, Li KW, Sciubba DM, Suk I, Wolinsky JP, Gokaslan ZL (2009) Posterior-only approach for total en bloc spondylectomy for malignant primary spinal neoplasms: anatomic considerations and operative nuances. Neurosurgery 65(6 Suppl):173–181 Liljenqvist UR, Halm HF, Link TM (1997) Pedicle screw instrumentation of the thoracic spine in idiopathic scoliosis. Spine 22:2239–2245 Suk SI, Lee CK, Kim WJ et al (1995) Segmental pedicle screw fixation in the treatment of thoracic idiopathic scoliosis. Spine 20:1399–1405 Hicks JM, Singla A, Shen FH et al (2010) Complications of pedicle screw fixation in scoliosis surgery: a systematic review. Spine 35:465–470 Blight AR, Decrescito V (1986) Morphometric analysis of experimental spinal cord injury in the cat: the relation of injury intensity to survival of myelinated axons. Neuroscience 19:321–341 Quencer RM, Bunge RP, Egnor M et al (1992) Acute traumatic central cord syndrome: MRI-pathological correlations. Neuroradiology 34:85–94 Donohue ML, Murtagh-Schaffer C, Basta J et al (2008) Pulsetrain stimulation for detecting medial malpositioning of thoracic pedicle screws. Spine 33:378–385 Falconer JC, Narayana PA, Bhattacharjee M, Liu SJ (1996) Characterization of an experimental spinal cord injury model using waveform and morphometric analysis. Spine 21:104–112
14. De Girolami U, Frosch MP, Tator CH (2002) Regional neuropathology diseases of the spinal cord and vertebral column. In: Graham DI, Lantos PL (eds) Greenfield’s neuropathology, 7th edn. Arnold, London, pp 1063–1101 15. Rivlin AS, Tator CH (1978) Effect of duration of acute spinal cord compression in a new acute cord injury model in the rat. Surg Neurol 10:38–43 16. Fehlings MG, Tator CH, Linden RD (1989) The relationships among the severity of spinal cord injury, motor and somatosensory evoked potentials and spinal cord blood flow. Electroencephalogr Clin Neurophysiol 74:241–259 17. Cao Q, Zhang YP, Iannotti C et al (2005) Functional and electrophysiological changes after graded traumatic spinal cord injury in adult rat. Exp Neurol 191(suppl 1):3–16 18. Scheff SW, Rabchevsky AG, Fugaccia I et al (2003) Experimental modeling of spinal cord injury: characterization of a force-defined injury device. J Neurotrauma 20:179–193 19. Poon PC, Gupta D, Shoichet MS, Tator CH (2007) Clip compression model is useful for thoracic spinal cord injuries. Histologic and functional correlates. Spine 32:2853–2855 20. Noble LJ, Wrathall JR (1989) Correlative analysis of lesion development and functional status after graded spinal cord contusive injuries in the rat. Exp Neurol 103:34–40 21. Carlson GD, Gorden CD, Oliff HS, Pillai JJ, LaManna JC (2003) Sustained spinal cord compression: part I: time-dependent effect on long-term pathophysiology. J Bone Joint Surg Am 85-A:86–94 ˚ , Sundstro¨m E (1997) Clip compression 22. von Euler M, Seiger A injury in the spinal cord: a correlative study of neurological and morphological alterations. Exp Neurol 145:502–510
123
