Thoracic myelopathy secondary to ossification of the spinal ligament

Transcription

1 J Neurosurg 66:11-18, 1987 Thoracic myelopathy secondary to ossification of the spinal ligament KAZUO YONENOBU, M.D., D.Ms., SOHE EBARA, M.D., KEJU FUJWARA, M.D., KAZUO YAMASH TA, M.D., KERO ONO, M.D., D.Ms., TOMO YAMAMOTO, M.D., NORMASA HARADA, M.D., HROSH OGNO, M.D., AND SHNZABURO OJMA, M.D. Departments of Orthopaedic Surgery, Osaka University Medical School, Osaka Koseinenkin Hospital, Hoshigaoka Koseinenkin Hospital, and Sumitomo Hospital, Osaka, Japan L-" The authors describe their experience with 26 cases of thoracic myelopathy secondary to hypertrophic ossification of the spinal ligament (posterior longitudinal ligament and/or ligamentum flavum). The clinical manifestations of this condition and results of its surgical treatment are described. The commonest symptoms were numbness or tingling in the legs and feet and gait disturbance. Most of the patients with involvement of the upper thoracic spine showed typical features of thoracic myelopathy: that is, sensory and motor deficits in both the trunk and lower extremities, sphincter disturbance, and exaggerated tendon reflexes. Several patients with involvement of the thoracolumbar junction presented with atypical symptoms of thoracic myelopathy and were sometimes misdiagnosed and treated inappropriately. Surgical treatment, particularly laminectomy, was not always successful. nconsistencies in the surgical outcome were caused by either operative complications or reversal of the initial improvement during the follow-up period. The results of anterior surgery for the condition were more favorable; however, use of this procedure was rarely indicated. KEY WORDS 9 thoracic myelopathy 9 ligamentum flavum 9 anterior decom posterior longitudinal ligament 9 ankylosing spinal hyperostosis 9 laminectomy ~ ligament ossification V AROUS diseases are known to produce hyperostosis or ossification of the ligament in the spine. ~2 One of these, ankylosing spinal hyperostosis, is a well-known condition that occurs worldwide. t was described by Forestier and Rotes-Querol 2 ~ n Japan, patients with ankylosing spinal hyperostosis may also develop ossification of the posterior longitudinal ligament (OPLL) in the cervical spine, with serious accompanying neurological complications. Techniques for diagnosis and treatment of OPLL in the cervical spine have been evaluated in a nationwide study by Japanese surgeons?,l~ The thoracic and lumbar portions of the spinal cord are also susceptible to com by hypertrophic ossification of the spinal ligament in the thoracic spinal canal. Heretofore, hypertrophic ossification of the ligamentum flavum (OLF) and OPLL have been known to be causes of thoracic myelopathy, particularly in Japan) '6-8 Unlike OPLL in the cervical spine, however, thoracic myelopathy due to ligament ossification may be overlooked, misdiagnosed, or treated inappropriately. This is mainly because of lack of knowledge of this condition, particularly outside Japan.~'9 For this reason we present our recent experience with this disorder and describe the clinical aspects, methods of diagnosis, and surgical treatment in a series of patients with this condition. Clinical Material and Methods Summary of Cases Between 1976 and 198, 26 patients with OPLL and/ or OLF of the thoracic spine were treated surgically in our special clinic for spinal disorders (Table 1). Men were afflicted predominantly by OLF, whereas a combination of OPLL and OLF occurred more frequently in women. The patients ranged in age from 8 to 67 years, but the majority were in their forties and fifties (Table 2). n general, OPLL occurred in the upper and middle thoracic spine and OLF at the thoracolumbar junction (Fig. 1). Thirteen of the 26 patients also had ankylosing spinal hyperostosis (Fig. 2), a disease frequently found in cases of OPLL of the cervical spine.,~z Six of our patients also had OPLL of the cervical spine, suggesting that ossification of the spinal ligaments has a common underlying etiology. Seven patients had J. Neurosurg. / Volume 66/April,

2 K. Yonenobu, et al. T1 T1 "--'--' s 2 OLF E OPLL r-- 1'o OPLL OLF 12 L OPLL & Total OLF Cases women 8 1 men total cases * OPLL = ossification of the posterior longitudinal ligament; OLF = ossification of the ligamentum flavum. Group 11 2 TABLE 1 Distribution of the 26 pateints in this series by sex and diagnosis*,'~ L 2 no. of level FG. 1. Graph showing the vertebral distribution of ossification of the posterior longitudinal ligament (OPLL) and ossification of the ligamentum flavum (OLF) among the 26 cases in this series. TABLE 2 Summary o f clinical data in the 26 cases in this series Factor age at surgery (yrs) duration of symptoms (mos) severity of myelopathy score* follow-up period (mos) Average Range * Scoring was performed using the Japanese Orthopaedic Association system (see Table ). A score of 1 l is normal. diabetes mellitus in c o m b i n a t i o n with the spinal lesion, although the etiological relationship between these two conditions has n o t been clarified. Diagnostic Procedures A detailed neurological examination was performed in each case, after which O P L L or O L F was identified by means o f x-ray imaging techniques such as plain lateral r o e n t g e n o g r a p h y and tomography. W h e n plain roentgenography could not delineate the lesion in the upper thoracic region or at the t h o r a c o l u m b a r junction, t o m o g r a p h y was useful in detecting both O P L L and OLF. The distinguishing features included either a thin strip o f increased density behind the vertebral b o d y in cases o f O P L L (Fig. ) or a beak-like projection into the intervertebral f o r a m e n in cases o f O L F (Fig. ). The presence o f ankylosing spinal hyperostosis in the t h o r a c o l u m b a r area m a y also indicate O P L L or O L F (Fig. 2). Neither the size o f the ossification n o r the canal diameter on plain radiograms or t o m o g r a m s was sufficient to delineate the s y m p t o m - p r o d u c i n g lesions, making m y e l o g r a p h y an indispensable study. C o m p u t e r i z e d t o m o g r a p h y with or without metrizamide was also helpf u l i n visualizing spinal cord i m p i n g e m e n t in planning the surgical a p p r o a c h (Figs. and 7). FG. 2. X-ray film showing ossification of the posterior longitudinal ligament in the upper thoracic spine and hyperostosis of the anterior aspect of the vertebral body, as well as ossification of the ligamentum flavum at the thoracolumbar junction. 12 Surgical Treatment Once thoracic m y e l o p a t h y developed, conservative treatment (including application o f a Jewett-type spinal brace a n d / o r d i p h o s p h o n a t e administration for prevenj. Neurosurg. / Volume 6 6 / A p r i l, 1987

3 Spinal ligament ossification and thoracic myelopathy FG.. Tomogram showing a shadow with the density of bone indicating ossification of the posterior longitudinal ligament behind the vertebral body from T-7 to T-10. FG.. Plain x-ray film (left) and tomogram (right) showing ossification of the ligamentum tlavum. This condition is more clearly seen in the tomogram. Note the beak-like ossified ligament in the foramen. TABLE Japanese Orthopaedic Association scoring system for assessment of dysfunction due to myelopathy Score Neurological Status motor dysfunction of lower extremities 0 unable to walk 1 able to walk on fiat floor with walking aid 2 able to walk up &/or down stairs with handrail lack of stability & smooth reciprocation of gait no dysfunction sensory deficit in lower extremities 0 severe sensory loss or pain 1 mild sensory loss 2 no deficit sensory deficit in trunk 0 severe sensory loss or pain 1 mild sensory loss 2 no deficit sphincter dysfunction 0 unable to void marked difficulty in micturition 2 minor difficulty in micturition no dysfunction FG.. Metrizamide-enhanced computerized tomography scan demonstrating hypertrophic ossification of the ligamenta flava in the spinal canal (arrows). tion of calcification) was not efficacious in preventing disease progression. When it became difficult or impossible for patients to walk, as determined by a score of 2 or less according to the Japanese Orthopaedic Association (JOA) scoring system (Table ), surgical treatment was indicated. nitially, we used posterior decom to treat both OPLL and OLF. Posterior decom was achieved in this series with a wide and extensive laminectomy that included resection of the laminae and half of the facets throughout the compressed area. Great care was taken not to damage the cord, because the dura was frequently adherent to the ossified ligaments. Because of the unfavorable results with laminectomy (described below in detail), anterior decom has recently been introduced for treating this condition. However, OPLL and OLF were accessible by the anterior approach at only three vertebral segments. The vertebral body and pedicle were exposed via a transpleural approach, and, after pediclectomy with a diamond airtome, the anterior wall of the spinal canal was exposed. Next, the OPLL was resected through a J. Neurosurg. / Volume 66/April,

4 K. Yonenobu, et al. FG. 6. Diagram showing resection of ossified posterior longitudinal ligament (shaded section) by means of the bilateral approach. First, through standard transpleural approach (as described in the text) ossified ligaments are removed and the spine is fused using a bone graft. Next, the rest of the ossified ligament was removed via an extrapleural approach from the opposite side. FG. 7. Preoperative (left) and postoperative (right) metrizamide-enhanced computerized tomography scans showing the spinal canal before and after resection of the ossified posterior longitudinal ligaments via the bilateral anterior approach. The spinal canal is enlarged postoperatively. subtotal corpectomy (Fig. 6). However, it was found that the OPLL on the side opposite to the thoracotomy was extremely difficult to resect. An anterior strut graft from the rib or ileum was used for spinal fusion. After weeks of bed rest, the patients were permitted to walk in a body cast. Two months after surgery the body cast was removed and a Jewett- or frame-type brace was applied until stability of the graft was confirmed. For lesions extending from T-1 to T-, a sternumsplitting approach was used for anterior decom and spinal fusion. For cases where a unilateral approach was not possible, we devised a bilateral approach to permit complete OPLL excision. Even if impingement on the spinal cord was caused by OLF, this anterior two-sided approach was used (Fig. 7), provided that the lesion was limited to three segments. 1 J. Neurosurg. / Volume 66/April, 1987

5 Spinal ligament ossification and thoracic myelopathy TABLE Symptoms and signs in the 26 patients in the series according to level of myelopathy* Symptoms & Signs Group 1 Group 2 Group no. of cases pain in leg or foot 6 (2.9) (0) 2 (.) sense of constriction in chest or (28.6) 2 (.) 0 abdomen numbness in leg or foot 1 (100) 6 (100) 6 (100) sensory loss trunk 1 (92.9) (0) 1 (16.7) thigh 1 (100) (0) 1 (16.7) leg 1 (100) (0) (8.) foot 1 (100) (66.7) (8.) hyperreflexia 1 (100) 6 (100) (0) motor weakness 11 (78.6) (66.7) 6 (100) gait disturbance 1 (100) 6 (100) 6 (100) sphincter disturbance 7 (0) (8.) 2 (.) * Group 1 patients had involvement at T9-10 or higher, Group 2 had involvement at T10-11, and Group had involvement at T 1-12 and L-1. Numbers in parentheses indicate percentages of the total number of patients in each group. Clinical Findings Results Onset of the symptoms was insidious and disease progression was very slow. Usually the initial symptoms were numbness or tingling dysesthesia in the soles of the feet and/or legs (12 of 26 patients) and sometimes difficulty in walking (nine of 26 patients). A stiff spine or back pain was not a common complaint. All of the patients eventually felt numbness or tingling dysesthesia of the lower extremities, especially in the lateral aspects of the legs and solar surfaces. Gait disturbance was commonly noted. Six patients were bedridden and six were able to walk only with aid; the rest were able to walk without aid but needed a handrail when going up or down stairs. Motor weakness was a common manifestation and was frequently recognizable in the long extensor muscle of the great toe and in the anterior tibial muscle. The hip girdle muscles, such as the abductor and psoas, were frequently weak. However, the degree of motor weakness was not severe in most patients, and 6.% of the patients had good (Manual Muscle Test Grade ) motor power in these muscles. The distinctive features of the manifestations upon admission to the hospital became clearer when the patients were assigned to one of three groups according to the level of involvement as determined by myelography (Table ). When the T9-10 level or higher was involved (Group 1, 1 patients), the manifestations were those of typical thoracic myelopathy. Loss of sensation extended one to two segments below the level of involvement with distal predominance. Severe sensory deficits such as analgesia or anesthesia were rare. Group 2 included six patients with involvement of the T level. Three of these patients showed hyperreflexia but the distribution of sensory disturbances did not correspond to the level of involvement. n two patients, sensory deficits were not detectable. These patients were initially diagnosed as having a motor neuronal disease. n Group 2 bladder sphincter disturbances were more common than in the other two groups. Most of the patients complained of hesitation on voiding. Pain was a frequent complaint, with pain in the thighs and legs often becoming intense while walking, similar to the symptoms of intermittent claudication. A sense of constriction in the chest or abdomen was described by Group 1 and 2 patients but was not accompanied by detectable hyperalgesia or severe sensory deficits in the trunk; thus, nerve root involvement was not suggested by these symptoms. Unlike patients with lesions at the upper thoracic level, three patients with involvement of T and L-1 (Group, six patients) showed a flaccid type of paraparesis. n this group, motor weakness in the leg muscles was consistently seen. Sensory deficits occurred in the lateral aspect of the legs and feet. Operative Outcome Results of surgery were assessed based on a recovery scale, which was calculated by the following equation: postop score - preop score recovery scale (%) = x [full score] - preop score According to this formula, the patient's postoperative neurological condition can be compared to the preoperative condition. The results were ranked as good (more than 0% improvement), fair (9% to 10% improvement), unchanged (9% to 0% improvement), and worse (less than 0% improvement). The scale is an integration of the score for each item in the scoring system proposed by the JOA (Table ). The preoperative, maximum, and final scores give the patient's functional profile before surgery, at the maximally improved stage, and at the final visit, respectively. Table summarizes the operative results. Nineteen of the 26 patients were treated initially by posterior decom. Four of these underwent additional anterior decom and three had additional posterior decom because of poor results after the initial surgery or for late regression. The other seven patients were initially treated by anterior decom, and two subsequently underwent posterior decom. Of the 19 patients treated initially with posterior decom, good results were obtained in six patients (1.6%). However, in four patients paraplegia developed immediately after surgery (Table 6) because the already affected spinal cord was damaged further during surgery or in the convalescent period. At the final follow-up examination, four patients were ranked as having a good outcome while the neurological condition in an additional four patients was worse (Table ). Cause of deterioration in the patients with late regression is summarized in Table 7 and includes kyphotic deformity and progressive ossification at other levels. Extensive laminectomy in the thoracolumbar J. Neurosurg. / Volume 66/April, fi

6 K. Yonenobu, et al. TABLE Results of surgery in the 26 cases in this series Surgery Operative Result* No. of Cases Good Fair changed Un- Worsened initial surgical procedure posterior decom maximally improved stage final examination anterior decom maximally improved stage final examination 0 1 second surgical procedure posterior decom: final examination anterior decom: final examination * For the method of calculating results see text. region seemed to be associated more frequently with kyphotic deformities than it was in other regions. Anterior surgery improved symptoms in most cases. nitially, all patients undergoing anterior decom benefited from surgery (Table ), and no patients showed marked progression of kyphosis. One month after the initial surgery one patient underwent an additional posterior decom to obtain further improvement, but his neurological condition was not ameliorated. Another patient became worse 8 months after surgery because of progression of OPLL of the cervical and thoracic spine (Case 7, Table 7). The incidence of complications during and after anterior decom was high (Table 6). Temporary neurological deterioration was encountered in three patients but all three eventually recovered. Their deterioration was caused by minor spinal cord damage that occurred during resection of the ossified ligaments. n three other patients cerebrospinal fluid continued to leak for to weeks after surgery, but the fistulas eventually closed without surgery. The results of a second surgical procedure were generally poor (Tables and 6). These patients all had severe myelopathy (average score of.2 by the JOA scale) before the second surgery. This may be why the results were poor. One patient in this group died due to bleeding after surgery. Only two patients improved slightly over the preoperative score. Discussion According to recently published reports, 6-8 hyperostosis of the ligament within the spinal canal is increasing in importance in Japan as one of the causes of paraplegia or paraparesis. However, a net increase in incidence of the condition has not been proven. More interest in spinal disease and the development of modern imaging technology are thought to have facilitated the discovery TABLE 6 ncidence of postoperative complications Complication nitial Surgery Second Surgery Posterior Anterior Posterior Anterior total cases no. (%) with complica (1.6) (71.) 1 (20.0) 2 (0.0) lions neurological deficit 1 l cerebrospinal fluid spinal instability dead of the condition. An accurate diagnosis of hyperostosis is also attributable to an expanded knowledge of conditions of the spine such as OPLL or ankylosing spinal hyperostosis. Upper thoracic hyperostosis typically includes thoracic myelopathy and is rather easily diagnosed as com myelopathy due to OPLL or OLF, provided that the physician is familiar with such conditions. On the other hand, the clinical features of patients with an ossifying lesion at the thoracolumbar junction are similar to those of central disc herniation or tumor of the cauda equina, and sometimes resemble neurological disorders such as Charcot-Marie-Tooth disease or lower motor neuronal diseases. Therefore, in the past the presence of this condition may have been missed or diagnosis delayed. The first step to a correct diagnosis is to conduct a thorough neurological examination. A sense of constriction of the trunk, girdle pain, decreased chest expansion due to a rigid thoracic cage, or a stiff back are often accompanying symptoms. The level of the lesion is indicated by the uppermost extent of motor and particularly sensory involvement, except in cases of thoracolumbar j unction lesions. Lateral plain roentgenograms or tomograms permit identification of the condition as either OLF or OPLL. Roentgenography should be performed with the focus at the presumed level of the lesion because visualization of the lesion may be hindered by overlapping of the shoulder girdle in the upper thoracic region where OPLL frequently develops or by the shadow of the liver at the thoracolumbar junction where OLF is often found. Myelography is necessary to delineate the symptom-producing lesion. During myelographic examination of patients suffering from flaccid paraparesis from an unknown cause, the thoracolumbar junction should be observed carefully. Computerized tomography scanning with or without metrizamide is very useful in visualizing the spinal canal and in assessing the relationship between the spinal cord and ossified ligaments. Magnetic resonance imaging (MR) might also be useful, especially at the upper thoracic level. However, we have not yet had enough experience with MR to determine its efficacy. Patients with paraplegia or paraparesis are treated by surgery as the last resort. However, despite the fact that fil 6 J. Neurosurg. / Volume 66/April, 1987

7 Spinal ligament ossification and thoracic myelopathy TABLE 7 Summary of seven cases with late regression following surgery for myelopathy* Time to Case nitial Surgery Preop Maximum Regression Score at Second Surgery Final Cause of Regression No. Score Score (mos) Regression Score 1 posterior decom posterior decom- 6 progression of OPLL of the upper segment 2 posterior decom anterior decom- 0 kyphotic deformity, progression of OPLL & OLF of the lower segment posterior decom- 6 l posterior decom- 8 recurrence of OLF posterior decom anterior decom- 0 kyphotic deformity, progression of posterior decom anterior decom- OPLL & OLF of the lower segment kyphotic deformity 6 posterior decom none trauma 7 anterior decom posterior decom progression ofopll & OLF of the upper segment * Scoring was performed according to the Japanese Orthopaedic Association system (see Table ). OPLL = ossification of the posterior longitudinal ligament; OLF = ossification of the ligamentum flavum. laminectomy has been used for treating the condition since Yamaguchi, et al., ~ reported a case in 1960, the appropriate procedure has yet to be established. The poor results of laminectomy in this series were mainly caused by neurological complications during and immediately after surgery and by late neurological regression. Four of 19 patients treated by posterior decom deteriorated postoperatively from undetermined causes. While a postoperative hematoma was probably one of the factors involved, advanced myelopathy and kyphotic deformity of the thoracic spine may have minimized the possible beneficial effect of posterior decom. Four patients became worse over a period of 6 to 86 months after surgery. Causes of the late deterioration proved to be progression of ossification within untreated segments, recurrence of the removed OLF, increased kyphotic deformity of the spine, or a combination of these reasons. Kyphotic deformity may relate to resection of the facets. We concluded that even a wide laminectomy did not assure an effective decom, since the thoracic spine is naturally kyphotic and it is unrealistic to expect the spinal cord to shift backward as a result of laminectomy as is the case with a lordotic cervical spine. Furthermore, in OPLL the spinal cord impingement occurs anteriorly in the canal. Therefore, we have changed the principal surgical approach from a posterior to an anterior technique. The results of anterior decom are better than those of posterior decom; however, the indication for anterior decom is different from that for posterior decom. The background data and follow-up period of both groups are not comparable (the average follow-up time was 2 months after ante- rior decom and 6 months after posterior decom). Therefore, it is difficult to compare the results of the two surgical approaches. So far, the results of anterior surgery seem to be more favorable than those of laminectomy, and spinal stability is restored at the same time. The extent of the decom by this procedure is, however, limited to three vertebral segments. References 1. Firooznia H, Benjamin VM, Pinto RS, et al: Calcification and ossification of posterior longitudinal ligament of spine. ts role in secondary narrowing of spinal canal and cord com. NY State J Med 82: , Forestier J, Rotes-Querol J: Senile ankylosing hyperostosis of the spine. Ann Rheum Dis 9:21-0, 190. Hattori A, Endoh H, Suzuki K, et al: Ossification of the thoracic ligamentum flavum with com of the spinal cord. A report of six cases. J Jpn Orthop Assoc 0: , Hirabayashi K, Miyakawa J, Satomi K, et al: Operative results and postoperative progression of ossification among patients with ossification of cervical posterior longitudinal ligament. Spine 6:-6, The nvestigation Committee on OPLL of the Japanese Ministry of Public Health and Welfare: The ossification of the posterior longitudinal ligament of the spine (OPLL). J Jpn Orthop Assoc :2-0, The nvestigation Committee on Ossification of Spinal Ligament: [nvestigation Committee Reports.] Tokyo: Japanese Ministry of Public Health and Welfare, 1982 (Jpn) 7. The nvestigation Committee on Ossification of Spinal Ligament: [nvestigation Committee Reports.] Tokyo: Japanese Ministry of Public Health and Welfare, 198 (Jpn) J. Neurosurg. / Volume 66/April,

8 K. Yonenobu, et al. 8. The nvestigation Committee on Ossification of Spinal Ligament: [nvestigation Committee Reports.] Tokyo: Japanese Ministry of Public Health and Welfare, 198 (Jpn) 9. Johnsson KE, Petersson H, Wollheim FA, et al: Diffuse idiopathic skeletal hyperostosis (DSH) causing spinal stenosls and sudden paraplegia. J Rheumatol 10:78-789, Onji Y, Akiyama H, Shimomura Y, et al: Posterior paravertebral ossification causing cervical myelopathy. A report of eighteen cases. J Bone Joint Surg (Am) 9: , Ono K, Ota H, Tada K, et al: Ossified posterior longitudinal ligament. A clinicopathological study. Spine 2: , Resnick D, Niwayama G: Diagnosis of Bone and Joint Disorders. With Emphasis on Articular Abnormalities. Philadelphia: WB Saunders, Yamaguchi H, Tamakake S, Hujita S: [A case of ossification of the ligamentum flavum with myelopathy.] Seikeigeka 11:91-96, 1960 (Jpn) Manuscript received July 11, Address reprint requests to: Kazuo Yonenobu, M.D., Department of Orthopaedic Surgery, Osaka University Medical School, Fukushima, Fukushima-ku, Osaka, Japan. 8 J. Neurosurg. / Volume 66/April, 1987