 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1998 The Journal of Bone and Joint Surgery, Inc. Dysfunction of the Spinal Cord During Spinal Arthrodesis for Scoliosis: Recommendations for Early Detection and Treatment. A Case Report*  , IOWA CITY, IOWAInvestigation performed at University of Iowa Hospitals and Clinics, Iowa City
Numerous factors can increase the risk of neurological injury during an operation for the correction of scoliosis; these include the presence of a congenital, severe, or rigid curve and the presence of abnormalities in the spinal cord27. Factors related to anesthesia, such as hemodilution and induced hypotension, may produce ischemia of the spinal cord16,15,27. Sublaminar wiring and derotation or distraction of the spinal column may also result in injury3,10,15,26. A careful clinical, neurological, and radiographic assessment is warranted for any patient who has scoliosis, especially when there are unusual findings; these findings may include the onset of scoliosis before the age of ten years, rapid progression of the curve, clinical symptoms such as back pain and headache, neurological symptoms or signs, and an atypical curve pattern2,23. Preoperative magnetic resonance imaging of the spine may demonstrate intraspinal abnormalities, such as syringomyelia associated with Chiari type-I malformation, spinal cord tumors, or diastematomyelia2,23. The presence of any of these conditions may increase the risk of morbidity during the operative correction of scoliosis, even after appropriate neurosurgical treatment20,21. Monitoring of somatosensory and motor evoked potentials has been associated with both false-positive results (observed when an abnormality in either the amplitude or the latency of the evoked potential, or the loss of potential, is detected despite normal findings on neurological examination) and false-negative results (observed when no abnormality in the amplitude or latency of the evoked potential is detected despite the presence of a true neurological deficit)6,17. The wake-up test alone is not reliable for assessment of the integrity of the spinal cord1,4,5,9,11,12,15. Many authors have recommended the use of both monitoring methods, when possible, during major operations on the spine6,7,13,14,17,19,24,25,28. We report the case of a twelve-year-old girl who had definitive changes in the function of the spinal cord, as indicated by variations in the somatosensory evoked potentials and the findings on the wake-up test during an operation for the correction of scoliosis. After appropriate treatment, the somatosensory evoked potentials and the results of the wake-up test returned to normal. This patient had atypical scoliosis and had had operative treatment for syringomyelia associated with Chiari type-I malformation before the spinal arthrodesis. To our knowledge, this is the first case to be reported in the English-language literature in which changes in the somatosensory evoked potentials and in the function of the spinal cord were reversed when appropriate treatment was given at the optimum time.
A twelve-year-old girl was first seen by us for the evaluation of scoliosis that had been detected during a school-screening program. The mother of the patient had noted a deformity of the back six months earlier. During this six-month period, the patient had intermittent headaches and back pain. Snoring and hoarseness had recently been noted. When she was seen by us, the girl reported no dysfunction of the bowel or bladder or any other health problems. She had not yet reached menarche. The family history was negative for scoliosis. Physical examination revealed a considerable thoracic prominence on the right side when the patient was in the forward-bending position. Neurovascular examination of the upper and lower extremities revealed normal findings. The quadriceps and Achilles-tendon reflexes were 3+ bilaterally. The abdominal reflexes were brisk. A posteroanterior radiograph of the spine, made with the patient standing, demonstrated considerable thoracolumbar scoliosis, with a Cobb angle8 of 90 degrees as measured from the cephalad margin of the body of the sixth thoracic vertebra to the caudad border of the body of the first lumbar vertebra (Fig. 1-A). A lateral radiograph demonstrated decreased thoracic kyphosis of 20 degrees and normal lumbar lordosis. The excursion of the iliac apophysis was quantitated as Risser sign22 1 to 2. Because of the large magnitude of the curve and the history of headache, back pain, snoring, and hoarseness, a magnetic resonance imaging study of the brain and spinal cord was performed.
The magnetic resonance images revealed a Chiari type-I malformation and a cervicothoracic syrinx. Approximately one month after the initial examination, the patient was managed with a laminectomy of the first cervical vertebra and decompression of the posterior fossa. Shunting from the fourth ventricle to the subarachnoid space and a duraplasty also was performed at the same time. The patient tolerated the procedure well, and there were no complications. The headaches, back pain, hoarseness, and snoring resolved by six weeks after the procedure. However, the scoliotic curve still had not improved by nine months postoperatively. Neurovascular function was intact, and radiographs of the spine demonstrated a curve of 98 degrees—an increase of 8 degrees compared with the measurement that had been made ten months earlier. The patient had not yet reached menarche. The findings on magnetic resonance imaging were interpreted as normal. Radiographs that were made one day before the operative treatment of the spinal deformity (approximately twelve months after the initial examination) demonstrated a scoliotic curve of 110 degrees (Fig. 1-B). Because of the magnitude and rigidity of the curve as demonstrated on side-bending radiographs, an anterior release and discectomy from the sixth thoracic to the second lumbar vertebra through a right thoracotomy, a posterior arthrodesis with segmental instrumentation from the second thoracic to the third lumbar vertebra, and a thoracoplasty to reduce a four-centimeter rib prominence was planned.
At the time of the operation, routine hypotensive anesthesia was induced in order to maintain a mean arterial pressure of sixty-five millimeters of mercury (8.66 kilopascals). The spine was exposed, through a right thoracotomy at the level of the sixth rib, from the sixth thoracic to the second lumbar vertebra. The segmental vessels were ligated in the midline at the same levels, and the discs and the end plates were removed. The disc spaces were loosely packed with morseled bone from the rib. Somatosensory evoked potentials remained normal during the anterior procedure. The patient was then turned prone for the posterior procedure. Dissection was carried out through a midline incision from the first thoracic to the third lumbar vertebra. The spine was exposed subperiosteally from the second thoracic to the third lumbar vertebra. A motor probe for the measurement of muscle action potentials16 was inserted between the fifth and sixth thoracic levels, but adequate signals for monitoring could not be obtained. A three-rod technique was used for correction of the scoliosis. First, a short rod was placed on the left side of the rigid segment, from the sixth to the eleventh thoracic vertebra. The rod was rotated to correct the alignment in the sagittal plane, and the rigid segment was distracted. Second, a long rod, with a claw configuration, was placed between the second and third thoracic levels superiorly and between the second and third lumbar levels inferiorly on the left (concave) side of the curve. Thirty minutes after the placement of this second rod, just before an appropriately sized and contoured rod was to be inserted on the right (convex) side of the curve, it was reported that the somatosensory evoked potentials could no longer be obtained (Fig. 2). A wake-up test was performed immediately; the patient was not able to move either lower extremity but could move both upper extremities and could respond to commands. Normotensive anesthesia then was induced, and the distraction on the short and long rods on the left side was decreased slightly. Monitoring of somatosensory evoked potentials showed normal signals on the right side but prolonged latency on the left side (Fig. 2). A second wake-up test revealed that the patient was able to move the right lower extremity but not the left. Subsequently, all screws in the spinal instrumentation were loosened so that all distraction forces were released. Monitoring of somatosensory evoked potentials revealed a return of all signals and normal latency on both sides (Fig. 2). A third wake-up test revealed normal findings. We decided to leave the hardware in place with only the amount of distraction that was necessary to hold all hooks securely and to maintain in situ correction. The contoured rod was then inserted on the right side of the curve without additional corrective forces being applied. Spinal cord monitoring continued to demonstrate normal sensory signals, and a fourth wake-up test revealed normal findings. Bone grafts were placed in the decorticated lateral gutters, and the wound was closed. Because we wanted to wake the patient as quickly as possible, no thoracoplasty was performed. On awakening in the recovery room, the patient was neurologically intact.
At the latest follow-up evaluation (two years after the operation), the patient had a good functional result as evidenced by good spinal balance, a full level of activity, and the absence of pain; however, she had only a fair cosmetic result because of a 65-degree residual curve and a three-centimeter residual rib prominence. Neurovascular examination revealed normal findings. Radiographs showed a stable fusion mass, with no loss of correction and a curve of 65 degrees (Fig. 1-C).
A detailed clinical, neurological, and radiographic assessment is necessary for any patient who has scoliosis. Atypical patterns of scoliosis usually necessitate a detailed workup. Magnetic resonance imaging is warranted when there are unusual findings, which may include the onset of scoliosis before the age of ten years, rapid progression of the curve, clinical symptoms such as back pain or headache, neurological symptoms or signs, or an atypical curve pattern2,23. Magnetic resonance imaging may reveal abnormalities of the spinal cord, such as syringomyelia associated with Chiari type-I malformation, tumors of the spinal cord, or diastematomyelia2,23. Most authors have agreed that these intraspinal malformations may cause scoliosis18. Moreover, these conditions may increase the risk of morbidity during the operative treatment of scoliosis20,21. Neurosurgical treatment of syringomyelia and Chiari type-I malformation may improve the clinical status and decrease the potential for neurological injury during later corrective operations on the spine. Muhonen et al. reported that neurosurgical procedures, such as decompression of the foramen magnum and shunting of the syrinx, may result in a considerable correction of scoliosis in patients who are younger than ten years old, even if the curve is greater than 40 degrees18. In the case of our patient, decompression of the posterior fossa and shunting of the syrinx resulted in clinical improvement as well as normal findings on magnetic resonance imaging studies. However, continued deterioration of the curve necessitated correction and spinal arthrodesis. The reasons for the altered function of the spinal cord during the corrective operation in this patient are unclear. We hypothesize that, despite neurosurgical treatment of the syrinx and the Chiari type-I malformation, there was increased fragility of the intraspinal vascularity that was further compounded by hypotensive anesthesia and distraction of the spine. Scar tissue in the subarachnoid space might also have had a tethering effect. However, we could find no evidence in the literature to support this hypothesis. Ben-David et al. reported false-negative results in association with both monitoring of somatosensory evoked potentials4 and the wake-up test6. Ginsburg et al.9 and Lesser et al.11 demonstrated that unchanged somatosensory evoked potentials do not guarantee normal neurological function. They also stated that the use of only one method of spinal cord monitoring is not sufficient. The belief that monitoring of somatosensory evoked potentials reflects the integrity of only the posterior portion of the spinal cord whereas the wake-up test indicates gross motor function has led many authors to recommend the use of both methods of monitoring during major operations, if possible6,7,13,14,17,19,24,25,28. We agree that the integrity of the anterior and posterior portions of the spinal cord can be reliably assessed only with use of a combined approach that involves the monitoring of both sensory and motor function or the monitoring of sensory function and the performance of the wake-up test. We recommend this combined approach because of the high risk of neurological compromise associated with the correction of neurogenic scoliosis, as demonstrated by the case of our patient. This combined approach minimizes the chance of a false-negative result. In our experience with more than 1000 patients who were monitored with this combined approach during the operative treatment of a spinal deformity, this patient was the only one for whom the findings of both methods of monitoring were in agreement, suggesting definite neurological impairment. Timely recognition of the problem, a reduction in the distraction forces on the spine, and the induction of normotensive anesthesia prevented permanent neurological impairment without complete removal of the instrumentation. Operative treatment of neurogenic scoliosis should be planned and performed carefully. Forceful correction of the curve, combined with induced hypotensive anesthesia, may be hazardous. The present study demonstrates the value of spinal cord monitoring in the detection of neurological injury in a high-risk patient.
*No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. No funds were received in support of this study.
  CiteULike  Connotea  Del.icio.us  Technorati What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Introduction
Case Report
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |