 |
 |
 |
|
| home | CME | help | feedback | subscriptions | archive |
|
© 1998 The Journal of Bone and Joint Surgery, Inc. Anterior Cervical Decompression and Arthrodesis for the Treatment of Cervical Spondylotic Myelopathy. Two to Seventeen-Year Follow-up*  , CLEVELAND, OHIOInvestigation performed at the Department of Orthopaedic Surgery, University Hospitals Spine Institute, Case Western Reserve University School of Medicine, Cleveland
We reviewed the cases of 108 patients with cervical spondylotic myelopathy who had been managed with anterior decompression and arthrodesis at our institution. Operative treatment consisted of anterior discectomy, partial corpectomy, or subtotal corpectomy at one level or more, followed by placement of autogenous bone graft from the iliac crest or the fibula. At the latest follow-up examination, thirty-eight of the eighty-two patients who had had a preoperative gait abnormality had a normal gait, thirty-three had an improvement in gait, six had no change, four had improvement and later deterioration, and one had a worse gait abnormality. Of the eighty-seven patients who had had a preoperative motor deficit, fifty-four had complete recovery; twenty-six, partial recovery; six, no change; and one had a worse deficit. The average grade according to the system of Nurick improved from 2.4 preoperatively to 1.2 (range, 0.0 to 5.0) postoperatively. A pseudarthrosis developed in sixteen patients, thirteen of whom had had a multilevel discectomy. Only one of thirty-eight arthrodeses that had been performed with use of a fibular strut graft was followed by a non-union. An unsatisfactory outcome with respect to pain was significantly associated with pseudarthrosis (p < 0.001). The development of complications other than non-union was associated with a history of one previous operative procedure or more (p = 0.005). Recurrent myelopathy was rare, but when it occurred it was associated with a pseudarthrosis or stenosis at a new level. The strongest predictive factor for recovery from myelopathy was the severity of the myelopathy before the operative intervention—that is, better preoperative neurological function was associated with a better neurological outcome. Anterior decompression and arthrodesis with autogenous bone-grafting can be performed safely, and is associated with a high rate of neurological recovery, functional improvement, and pain relief, in patients who have cervical spondylotic myelopathy.
Cervical spondylosis is a result of degenerative changes in the discs and the formation of chondro-osseous spurs, which ultimately can lead to compression of the spinal cord and myelopathy4,6,26,30. Clarke and Robinson8 demonstrated the progressive nature of untreated cervical spondylotic myelopathy, and operative intervention is indicated in many such patients to improve or at least to preserve the neurological condition. The present study is a review of the cases of 108 patients managed at one institution, over a nineteen-year period, with anterior decompression and autogenous bone-grafting at one level or more for the treatment of cervical spondylotic myelopathy. Our goal was to report the long-term clinical outcome, with regard to pain, neurological recovery, late deterioration, and complications, and to relate the preoperative variables, such as the age of the patient and the grade of the myelopathy according to the system of Nurick27, to the clinical and roentgenographic results.
One hundred and eight patients with cervical spondylotic myelopathy who had had anterior decompression and arthrodesis at our institution between 1974 and 1993 and had been followed for at least two years were included in the present review. Myelopathy was diagnosed if abnormal reflexes, such as clonus, a positive Babinski sign, or a positive Hoffmann sign, were evident on physical examination or if the patient had a demonstrable disturbance of gait and hyperactive reflexes. Patients who had hyperreflexia but no other signs were not included. Neuroroentgenographic evidence of compression of the cervical spinal cord from herniated disc material or osteophytes, or both, was necessary to confirm the diagnosis. Patients who had segmental ossification of the posterior longitudinal ligament were included, but those who had continuous ossification of the posterior ligament over several vertebral levels were not. Patients who had neck pain alone, radiculopathy alone, cervical stenosis without myelopathy, myelopathy caused by severe kyphosis, or a traumatic injury were excluded. The operative procedure consisted of a Robinson anterior cervical discectomy and arthrodesis with use of an autogenous tricortical iliac-crest bone graft; a partial corpectomy, with removal of the disc and a portion of each vertebral body cephalad and caudad to the disc, followed by placement of a tricortical iliac-crest bone graft; or a subtotal vertebrectomy, with removal of all but the lateral walls of the vertebral body back to the posterior longitudinal ligament at one level or more, followed by insertion of an autogenous iliac-crest bone graft or a fibular strut graft. The specific procedure and operative levels were chosen by the attending surgeon on the basis of the findings of the preoperative neuroroentgenographic studies. Two patients died of cardiopulmonary complications in the early postoperative period. Another patient, who had a follow-up examination nine months postoperatively and was interviewed by telephone four years postoperatively, declined to return for an office visit. All other patients were followed for at least two years (range, two to seventeen years; average, 4.5 years). All patients were examined by the operating surgeon (H. H. B., S. E. E., or M. J. B.), except for four who were unable to return because it was too far to travel; these four were examined by their local orthopaedist or neurologist. Sixty-nine of the 108 patients in the study were men and thirty-nine were women. At the time of the operation, the average age of the patients was fifty-eight years (range, twenty-seven to eighty-eight years). Only two patients were receiving Workers' Compensation. Sixteen patients had had a previous operation on the neck at another institution; one of these patients had had six previous operations, and two had had two previous operations. Two patients had a pseudarthrosis that had developed after a previous anterior arthrodesis. The duration between the onset of the symptoms and the index procedure was less than one month for six patients, one to twelve months for forty-five, between one and two years for twenty-one, and more than two years for thirty-six. Preoperative symptoms included neck pain (thirty-seven patients), pain in the neck and upper extremity (forty-eight), pain in the upper extremity alone (four), and myelopathy with no pain (nineteen). Pain was graded according to the classification of Robinson et al.33 before and after the operation. It was classified as mild if there was no restriction of activities and the patient only occasionally used anti-inflammatory medication for relief, as moderate if there was slight limitation of activities and the patient frequently used non-narcotic pain medication for relief, and as severe if there was regular restriction of activities and the patient used narcotic medication for pain relief. Eighty-nine patients had pain preoperatively; it was mild in twenty patients, moderate in thirty-six, and severe in thirty-three. Thirty-five patients reported subjective weakness of the upper extremities; thirteen, weakness of the lower extremities; and fifty-four, weakness of both the upper and the lower extremities. Eighty-one patients had sensory symptoms in the upper extremities, and twenty-three had symptoms in the lower extremities. Twenty-four patients noted no disturbance of gait or balance; thirty-five reported that there was a difference in gait or balance but the ability to walk or to change positions was minimally affected; twenty-three had a moderate-to-severe change in the ability to walk that caused them to take compensatory action, such as leaning against a wall or on furniture; thirteen patients used a cane; five patients used a walker; and eight were dependent on a wheelchair. On physical examination, eighty-seven patients demonstrated motor weakness. Forty-four patients had motor weakness of the upper extremity alone, which was unilateral in seventeen and bilateral in twenty-seven; five had motor weakness of the lower extremity alone, which was unilateral in three and bilateral in two; and thirty-eight had motor weakness of both the upper and the lower extremities. Sixty-six patients had involvement of more than one muscle group. On sensory examination, eighty-nine patients had sensory loss in the upper extremities, trunk, or lower extremities; most of these sensory findings involved multiple dermatomes or exhibited a non-dermatomal pattern. Proprioception was abnormal to testing in twenty-two patients. Seventy-one patients had a demonstrable disturbance of gait, and seventeen were quadriparetic. All 108 patients had hyperreflexia, ninety-one had a positive Hoffmann reflex, and seventy-six had either a positive Babinski reflex or clonus. Preoperatively, twenty-two patients had no functional impairment despite the symptoms and physical findings. Fifteen patients had mild impairment when performing strenuous labor or participating in recreational activities, twenty were not able to work or to perform vigorous activities at home but were able to perform the activities of daily living, sixteen had an impaired ability to perform the activities of daily living, twenty-four had an impaired ability to walk in the community, two were only able to walk about the house, and nine used a wheelchair. The system of Nurick27 was used to classify the severity of the disability associated with the myelopathy before the operation and at the long-term follow-up examination (Table I). Preoperatively, thirty-eight patients had grade-1 myelopathy; nineteen, grade-2; twenty-six, grade-3; seventeen, grade-4; and eight, grade-5. The average grade of myelopathy was 2.4.
Roentgenographic Analysis Myelography was performed in 103 patients, and the findings included a complete or nearly complete block (twenty-five patients), a transverse anterior bar defect (forty), a nerve-root cut-off (twenty-nine), and ventral extradural defects (thirty-nine). Compression at multiple levels of the spinal cord or nerve roots was demonstrated in most patients. Seventy-five patients had computer-assisted tomography after myelography, and the findings included moderate-to-severe deformation of the spinal cord (seventy patients), mild flattening of the cord (six), and segmental ossification of the posterior longitudinal ligament (thirteen). Compression of the spinal cord was attributed to disc herniation alone in four patients and to a combination of disc herniation and osseous compression in the remaining seventy-one patients. Magnetic resonance imaging was performed on sixty-six patients, and it was the only preoperative neuroroentgenographic study made for five of them. The findings generally paralleled those of the myelographic studies, although an atrophic cord was revealed by one magnetic resonance imaging study and parenchymal signal changes consistent with myelomalacia were demonstrated by another. The cross-sectional area of the spinal cord at the level of maximum compression was measured preoperatively, with use of computer-assisted measurement of the spinal cord from the roentgenographic study (Bioquant IVa system; R and M Biometrics, Nashville, Tennessee), in the patients who had had a magnetic resonance imaging study or a computed tomography myelogram; the measurement could be obtained in sixty-three patients. The area of the spinal cord at the level of maximum compression was analyzed with respect to the preoperative and postoperative grade of myelopathy according to the scale of Nurick27 to determine if it had predictive value. At the time of the latest follow-up, the severity of persistent pain as well as the functional status were assessed. A complete neurological examination was performed by one of us (S. E. E., H. H. B., or M. J. B.), and lateral roentgenograms were made. The criteria for fusion were trabecular bridging of bone as well as absence of motion as seen on roentgenograms made with the spine in flexion and extension.
Indications for Operative Intervention
Operative Technique
If disc material, osteophytes, or an ossified posterior longitudinal ligament were present posterior to the vertebral body, then subtotal corpectomy was performed to decompress the spinal canal adequately and safely (Figs. 2-A, 2-B, and 2-C). First, the material was removed from the disc spaces with currettage and then the middle section of the body was excised so that only lateral shells of bone remained. Care was taken not to remove bone too far laterally in order to avoid injury of the vertebral artery37. The trough made in the vertebra was usually sixteen to eighteen millimeters in width, which provided decompression of the spinal canal and enough space for placement of the strut graft. Diamond burrs were used to thin the posterior shell of the vertebral body back to the posterior longitudinal ligament, after which it could be elevated off the ligament with use of tiny curets. If the posterior longitudinal ligament was intact and had not ossified, we did not remove it. The docking sites of the graft into the end vertebrae were sculpted with a burr to expose bleeding subchondral or cancellous bone. Anterior and posterior lips were fashioned to prevent migration of the graft. Skeletal traction was used for placement of the strut graft, and traction was increased when the graft was tapped into place. The strut graft was centered and then slightly countersunk within the docking site of each end vertebra. In a one-level subtotal corpectomy, an iliac-crest strut graft was used; in a two-level procedure, the surgeon chose a strut graft from either the iliac crest or the fibula; and in a three or four-level corpectomy, a fibular graft was used (Figs. 2-A, 2-B, and 2-C). All grafts were autogenous bone.
Intraoperative monitoring of the spinal cord with cortical evoked potentials was used for most of the corpectomy procedures. No patient had internal fixation. Postoperatively, a soft rubber drain was used and the head of the bed was elevated 20 to 30 degrees. All patients who had had a discectomy and an arthrodesis or a partial corpectomy and an arthrodesis wore a rigid head-cervical-thoracic orthosis postoperatively. Of the patients who had had a subtotal corpectomy, forty-four wore this type of orthosis and thirteen wore a halo vest. The choice of a brace or a halo vest was made on the basis of the bone quality, the intraoperative stability of the graft, and the personal preference of the attending surgeon. Forty-five of the 108 patients had a standard Robinson anterior cervical discectomy and arthrodesis. One level was involved in fourteen patients; two levels, in fifteen; and three levels, in sixteen. Two patients had a partial corpectomy at one level, and four had a partial corpectomy at one level and a discectomy at one or two additional levels. Fifty-five patients had a subtotal corpectomy; it was performed at one level in eleven patients, at two levels in twenty-two, at three levels in twenty-one, and at four levels in one. Two other patients had a corpectomy at one level and a discectomy at another, and both had insertion of bone graft from the iliac crest. A strut graft from the iliac crest was used in nineteen patients who had a subtotal corpectomy, and a fibular strut graft was used in thirty-eight.
Statistical Analysis
Two patients died in the perioperative period; one had a myocardial infarction one week postoperatively, and the other had pulmonary failure from pneumonia approximately six weeks postoperatively. The remaining 106 patients were followed for at least two years. The twenty-four patients who had had no gait abnormality preoperatively had no change in gait postoperatively. Of the eighty-two patients who had had a preoperative gait abnormality, thirty-eight (46 per cent) had a normal gait, thirty-three (40 per cent) had an improvement in gait, six (7 per cent) had no change, four (5 per cent) had an initial improvement and later deterioration, and one (1 per cent) had a worse gait. The nineteen patients who had had no motor weakness preoperatively had none postoperatively. Of the eighty-seven patients who had had a preoperative motor deficit, fifty-four (62 per cent) had complete motor recovery, twenty-six (30 per cent) had partial recovery, six (7 per cent) had no change, and one (1 per cent) had a worse deficit. Weakness from nerve-root compression at another level developed later in two patients. Of the eighty-nine patients who had had a sensory deficit preoperatively, forty-three had complete recovery, thirty-five had improved sensation, ten had no change, and one lost sensation. A pseudarthrosis developed in sixteen patients. Thirteen of them had had an anterior cervical discectomy and arthrodesis (six had had a two-level and seven had had a three-level procedure). Two patients had non-union after a one-level subtotal vertebrectomy with insertion of a strut graft from the iliac crest. Of the thirty-eight patients who had a fibular strut graft, only one had a non-union; all others had a successful arthrodesis. Of the sixteen patients who had a pseudarthrosis, ten had neck pain and three had recurrent myelopathy that was attributed to the pseudarthrosis. Five of the ten patients who had a painful non-union had enough symptoms to warrant a revision operation. Moderate-to-severe pain at the time of follow-up was highly associated with the presence of a pseudarthrosis (p < 0.001). Chi-square analysis was used to compare the risks of pseudarthrosis associated with the different types of grafts. No significant difference was found, with the numbers available, between the rate of non-union associated with Robinson-type grafts and that associated with strut grafts from the iliac crest (p > 0.10). The rate of union associated with fibular strut grafts (97 per cent), however, was significantly higher than that associated with Robinson-type grafts (71 per cent) (p < 0.002). The rate of non-union in the patients who had had a multilevel anterior cervical discectomy and arthrodesis was 42 per cent compared with 0 per cent in those who had had a one-level procedure (p < 0.002). There were few cigarette smokers in our series, and smoking was not found to be associated with the development of a pseudarthrosis. Age was negatively correlated with the occurrence of a pseudarthrosis (r = -0.24, p < 0.05)—that is, older patients were more likely to have a solid fusion. Of the eighty-nine patients who had had preoperative pain, sixty-one (69 per cent) had no pain, sixteen (18 per cent) had mild neck pain, five (6 per cent) had moderate neck pain, and one (1 per cent) had moderate radicular pain. Pain developed after a postoperative infection secondary to extrusion of the iliac-crest graft and esophageal perforation in one patient who had had no preoperative pain. Of the 106 patients who had been followed for at least two years, seventy-one had no functional impairment, fifteen had mild impairment of the ability to perform strenuous labor or to participate in sports, eight had moderate impairment (they could perform the activities of daily living but could not work), two had impairment of the ability to perform the activities of daily living, six had impairment of the ability to walk in the community, two were able to walk only about the house, and two were unable to walk. At the latest follow-up examination, the average grade according to the system of Nurick27 was 1.2 (range, 0.0 to 5.0). Age was the only factor analyzed that was found to be associated with the preoperative grade according to the system of Nurick27—that is, older patients tended to have more severe myelopathy before the operation. The strongest predictor of the postoperative grade27 of the myelopathy was the preoperative grade: better neurological function before the operation was correlated with a better neurological outcome postoperatively (r = 0.64, p < 0.001). In addition, after controlling for the preoperative grade of the myelopathy, we found that women had less improvement than men in terms of the postoperative grade (partial r = 0.31, p < 0.01). The myelopathy recurred in five patients. Three of them had a pseudarthrosis with resultant formation of osteochondral spurs and recurrent compression of the spinal cord. One of the patients declined additional operative intervention. The other two had a revision operation (a posterior procedure in one and an anterior procedure in the other), with subsequent clinical improvement. Stenosis of the spinal canal at new levels with recurrent myelopathy developed in two patients; both had another anterior decompression and arthrodesis. One of these two patients had improvement in motor strength and function. The neurological deterioration was halted in the other patient, but there was little improvement with respect to the myelopathy. Six other patients needed an additional operative procedure for the treatment of disc herniation or spondylosis at adjacent levels, and one patient needed posterior stabilization because of subluxation caudad to the site of a long fusion.
Analysis of the Diameters of the Spinal Cord and the Canal In the sixty-three patients for whom measurements were available, the average cross-sectional area of the spinal cord was 34.6 square millimeters, with a wide range from 8.9 to 110.9 square millimeters (median, 32.8 square millimeters). The area of the spinal cord also was not found to be associated with the preoperative or postoperative severity of the myelopathy.
Complications Postoperative obstruction of the upper airway due to edema10 required reintubation in three patients. One of these patients had pneumonia with subsequent respiratory failure and died approximately six weeks postoperatively. Four patients had an intraoperative leak of cerebrospinal fluid, which was successfully repaired with a fascial patch and lumbar drainage of the cerebrospinal fluid36 in three patients. The remaining patient needed coverage with a muscle flap in order to seal the leak. Six patients had complications related to the graft. Four of them had displacement of the strut graft necessitating operative revision. One patient had partial displacement of the strut graft, which was treated by changing the type of immobilization to a halo vest, and one had collapse of an osteoporotic iliac-crest strut graft, which was treated with operative revision to a fibular bone graft. Erosion of the esophagus and postoperative infection developed in one of the patients who had displacement of an iliac-crest strut graft. The patient was managed successfully with operative débridement and replacement of the graft, nasogastric suction, and systemic antibiotic therapy. Two of the six patients who had a complication related to the graft were initially managed with halo immobilization. None of the Robinson-type grafts collapsed or became displaced. A hematoma developed in two patients postoperatively; one hematoma drained spontaneously, and the other was treated with operative drainage, with no adverse sequelae. Of the twelve patients who had at least one major complication, five had had operative intervention before the index procedure (p = 0.005). Of the six patients who had a complication related specifically to the graft, four had had a previous laminectomy (p < 0.001). Minor complications included a hematoma at the donor site of the graft, which resolved; a fracture at the donor site of an iliac-crest graft, which was treated symptomatically; a delayed tibial stress fracture (after bone graft was obtained from the fibula), which healed after application of a splint12; and an infection of the urinary tract in one patient and an infection at the site of a halo pin in another, which were both treated with antibiotic therapy.
Operative treatment of cervical myelopathy has focused on decompression of the spinal cord to halt neurological deterioration and to promote recovery. Both anterior and posterior operative techniques have been used to decompress the canal, with varying rates of success. Theoretically, posterior laminectomy is technically less demanding and requires no healing of a bone graft13. Problems with laminectomy have included postoperative instability, inadequate decompression of the spinal cord, persistent pain, and progression of the neurological deficit17,23,25,39. Laminaplasty has been performed as an alternative to laminectomy in an attempt to avoid the destabilizing effects associated with removal of the lamina19,22. Laminaplasty has limitations, however, as it is an indirect method of decompression, it results in some loss of motion, it is not recommended for patients who have cervical kyphosis24, and it may not treat axial neck pain successfully23. During the past twenty-five years, we have preferred to approach the spine anteriorly when treating cervical myelopathy. Anterior decompression allows direct removal of the compressive abnormality, with stabilization obtained by anterior arthrodesis. The disadvantages include the technical challenge of the procedure, the need for healing of a graft, and potential problems at adjacent levels. One of us2, in 1977, described seventeen patients who had moderate-to-severe myelopathy that was treated with anterior discectomy and arthrodesis. The patients had good neurological recovery, and many regained the ability to walk about the house or the community. However, the duration of follow-up ranged from six months to six years, and pain relief was not described in detail. Ebersold et al.9, in a study of thirty-three patients, noted very good early results after anterior discectomy and arthrodesis; some neurological deterioration occurred over time in six patients, but several of them had an additional diagnosis of multiple sclerosis, arteriovenous fistula, or poliomyelitis. No data on the outcome with respect to pain was described. Other authors have reported the results of subtotal corpectomy and strut-grafting for the treatment of cervical myelopathy1,2,18,29,34,39,41. Bernard and Whitecloud1 reported good functional recovery and no non-unions after use of fibular grafts with a dovetail technique. Yonenobu et al.39 and Okada et al.29 reported the results of subtotal corpectomy and anterior arthrodesis for the treatment of spondylotic myelopathy. In both studies, there was a high rate of improvement in neurological function and few patients who had late deterioration. None of these studies, however, addressed the outcome with respect to pain. In a study by Saunders et al.34, in which forty patients who had spondylotic myelopathy were managed with subtotal corpectomy and use of a strut graft from the iliac crest or the fibula, the rate of neurological recovery was high and only two patients had deterioration over time. The rate of complications was 48 per cent, but it included minor, treatable complications such as hyperventilation and hyponatremia. Our report is a retrospective review of the long-term results of anterior decompression and arthrodesis, performed at one institution by three surgeons who used consistent techniques, in patients who had cervical spondylotic myelopathy. As other investigators have done15,20,28,38, we attempted to identify the roentgenographic and clinical predictors of the preoperative and postoperative severity of the myelopathy. We noted a strong association between the severity of the myelopathy preoperatively and that at the latest follow-up examination, an observation that was in agreement with the findings of Saunders et al.34. To our knowledge, gender has not been reported to be associated with the outcome of treatment of myelopathy, but our data suggested that men have more improvement, even after we controlled for the preoperative rating of severity. We believe that the more salient point of our results is that most patients, male or female, have substantial postoperative improvement. The duration of symptoms; the sagittal area of the spinal canal at its narrowest point, as seen on plain roentgenograms; the Pavlov ratio31; and the preoperative area of the spinal cord, as measured on cross-sectional imaging studies, were not found to be associated with the severity of the myelopathy either preoperatively or postoperatively, with the numbers available. It is likely that plain roentgenograms do not reflect the true degree of compression of the spinal cord as soft-disc herniation and even osteophytes cannot be visualized accurately. We were surprised that the area of the spinal cord had no predictive value, but the result may have been influenced by the fact that cross-sectional imaging data were available for only sixty-three of the 108 patients. The improvement in motor strength, gait, and function as well as the reduction of sensory deficits was very high, with many patients returning to normal or nearly normal neurological function. One patient, however, had a catastrophic neurological outcome resulting in complete quadriplegia. The reason for the increased deficit was unclear, but it was believed to be related to intraoperative traction and hypotension occurring after the decompression, with resultant ischemia of the spinal cord14. Most of our patients had either axial or radicular pain in conjunction with the spondylotic myelopathy. Although there is little information in the literature regarding the outcome with respect to pain for such patients, it has been our experience that anterior stabilization of the spondylotic segments results in substantial relief of neck pain. The long-term results in our study demonstrated no or only mild pain in seventy-seven (87 per cent) of the eighty-nine patients who had had pain preoperatively. Pseudarthrosis was associated with more pain postoperatively, a finding that we also reported in an earlier study on anterior discectomy and arthrodesis for the treatment of cervical radiculopathy5. The risk factors for pseudarthrosis included the number of operative levels involved in an arthrodesis performed with use of a Robinson-type horseshoe-shaped bone graft33 and the source of the bone graft (the iliac crest or the fibula). We believe that our technique, in which the entire cross-sectional surface of the fibular graft is centered into the end vertebrae, provides maximum stability and surface area for healing; the rate of union of autogenous fibular grafts inserted in this manner has continued to be extremely high. A younger age is also associated with a higher risk of pseudarthrosis. We can only postulate that younger, healthier patients may be more active and have more mobility in the neck, which results in a slightly less stable environment for the bone graft, but we have no data to support this hypothesis. Six of the twelve patients who had a major complication had difficulties related to the strut graft. We believe that it is notable that four of the six patients in whom the strut graft dislodged or collapsed had had a previous multilevel cervical laminectomy. A compromised posterior column places an anterior strut graft at a substantially higher risk for displacement. A rigid two-poster brace was used for four of the six patients who had a complication associated with the graft, and two of the four patients had had a previous laminectomy. Although a halo vest provides better immobilization than a brace, it does not prevent axial loading or all complications related to the graft. Alternative methods to minimize dislodgment of the graft include use of a plate anteriorly21 or posterior stabilization; we have used both of these methods, but not in any of the patients in the present series. Osteoporosis was another important factor related to complications involving the strut graft in our patients. No bone-densitometry data were available; however, on the basis of the preoperative roentgenograms and the operative findings, we believe that four of the six patients who had a complication related to a strut graft had osteoporosis. If a patient has osteoporosis, there is some risk that the anterior part of the body of the inferior vertebra may fracture during axial loading of the strut graft when the patient assumes an upright posture. Some investigators have thought that a cervical fusion increases biomechanical stress at the remaining levels16 and perhaps accelerates degenerative changes7. The prevalence of symptomatic problems at adjacent disc levels in nine (8 per cent) of the 106 patients in our study is similar to that reported after anterior discectomy and arthrodesis procedures in patients who had cervical radiculopathy5. It is impossible to determine from our data whether this finding represents the normal progression of degenerative changes or an acceleration of the changes because of the fused segments; we believe that symptomatic problems at the adjacent disc levels are most likely a combination of the two processes, which become clinically important in a relatively small percentage of patients. The 5 per cent prevalence of recurrent myelopathy in our series was similar to, or better than, that reported in other studies1,9,29,34,35. Compression of the spinal cord and recurrent myelopathy developed in two of our patients who had spondylosis at an adjacent level and in three patients who had a pseudarthrosis. Persistent motion at a spondylotic segment can result in an increase or a recurrence of chondro-osseous spurs, which can lead to static or dynamic compromise of the spinal canal. We recommend that roentgenograms with the spine in flexion and extension, tomography, or computed tomography with spinal reconstruction be made in order to look for a pseudarthrosis. Magnetic resonance imaging studies or computed tomography myelograms should then be made in order to look for extrinsic or intrinsic conditions that could explain the recurrent myelopathy. In conclusion, our long-term study of the results of anterior decompression and arthrodesis for the treatment of cervical myelopathy demonstrated a high rate of improvement in neurological function or complete neurological recovery as well as substantial or complete relief of pain in most patients who had had preoperative pain. Functional improvement paralleled the neurological recovery. The most common major complications were related to the graft, and they occurred in patients who had had at least one previous operative procedure on the cervical spine. Multilevel anterior discectomy and arthrodesis procedures were associated with the highest rate of pseudarthrosis, and arthrodesis with an autogenous fibular strut graft was associated with the highest rate of success.
NOTE: The authors acknowledge the assistance of Timothy Pringle, M.D., in preparing the data regarding the cross-sectional area of the spinal cord.
*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  Facebook  Technorati  Twitter What's this?
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Top
Abstract
Introduction
| home | CME | help | feedback | subscriptions | archive | site index |
|
Stanford University Libraries' HighWire Press (TM) assists in the publication of JBJS Online. |