9. Klose U. In vivo proton spectroscopy in presence of eddy cur10.
11.
12.
13.
14.
15.
rents. Magn Resun Med 1990;14:26-30 Frahm J, Bruhn H, Gyngell ML, et al. Localized proton NMR spectroscopy in different regions of the human brain in vivo. Relaxation times and concentrations of cerebral metabolites. Magn Reson Med 1989;11:47-63 Petroff OA, Spencer DD, Alger JR, Pritchard JW. High-field proton magnetic resonance spectroscopy of human cerebrum obtained during surgery for epilepsy. Neurology 1989;39: 1197-2202 Medina L, Chi TL, DeVivo DC, Hilal SK. MR findings in patients with subacute necrotizing encephalomyelopathy (Leigh syndrome): Correlation with biochemical deficit. AJNR 1990; 11:379-384 Frahm J, Michaelis T, Merboldt K-D, et al. Localized NMR spectroscopy in vivo. NMR Biomed 1989;2:188-195 Birken DL, Oldendorf WH. N-acetyl-L-aspartic acid: a literature review of a compound prominent in ‘H-NMR spectroscopic studies of brain. Neurosci Biobehav Rev 1989;13:23-31 Tofts PS, Wray S. A critical asessment of methods of measuring metabolite concentrations by NMR spectroscopy. NMR Biomed 1988;l:l-10
Cerebrospinal Fluid Dynamics in the Tardive Cauda Equina Syndrome of Ankylosing Spondylitis Christian Confavreux, MD,” Jean-Paul Iarbre, MD, 1 Edouard Lejeune, MD,t Marc Sindou, MD,$ and Gilbert Aimard*
Typical cauda equina syndrome secondary to longstanding ankylosing spondylitis is reported in a 63-yearold man. Radionuclide cisternography demonstrated a resorption defect of cerebrospinal fluid in the enlarged lumbosacral dural sac. After transient symptomatic improvement with acetazolamide, a lumboperitoneal shunt was placed. T h e rate of cerebrospinal fluid, isotope resorption became normal. In the 5 years of follow-up, partial remission has been observed. Confavreux C, Larbre J-P, Lejeune E, Sindou M, Aimard G. Cerebrospinal fluid dynamics in the tardive cauda equina syndrome of ankylosing spondylitis. Ann Neurol 1991;29:221-223
Tardive cauda equina syndrome in ankylosing spondylitis (CES-AS) is a rare but well-known entity since From the Tlinique de Neurologie, HBpital Neurologique, and $Service de Neurochirurgie, HBpital Neurologique, Lyon, and +Service de Rhumatologie, HBpitd des Charpennes, Villeurbanne, France. Received Apr 10, 1990, and in revised form Jun 20 and Aug 8, 1990. Accepted for publication Aug 10, 1990. Address correspondence to Dr Confavreux, Clinique de Neurologie, HBpital Neurologique, 59 Boulevard Pinel, 69003 Lyon, France.
Babinski’s pioneer work [l}. To date, 69 cases have been reported 12-7}. Inflammation and fibrosis predominating o n dura matter and epidural space associated with enlargement of the caudal sac, narrowing of epidural space, adherence of dura matter to adjacent bony and ligamentous structures, and bony erosions of laminae and spinous processes are the basic radiological [Z, 81, surgical 13-5, 7, 9-11], and pathological C6, 121 features. The long interval separating the onset of ankylosing spondylitis (AS) and cauda equina syndrome (CES) and the common finding that AS has been burned out for several years when CES begins suggests that active inflammation is not directly involved in its pathogenesis. Alternatively, retractile fibrosis in the epidural space could lead t o a “between-trunk-andbark” variety of entrapment syndrome. We report a patient, with study of the cerebrospinal fluid (CSF) dynamics by isotopic techniques. A resorption defect of the isotope from the caudal sac was observed. Correction by lumboperitoneal shunting resulted in partial improvement of radicular pain.
Patient Report A 63-year-old, HLA-B27 positive, graduate engineer suffered from a relapsing form of AS since the age of 23 years. Sacroiliac joints and spine became fused with pronounced kyphosis. In 1956, vertebral osteotomy with posterior osseous graft was performed at the T12-L1 level for correction of the spinal deformation. During the subsequent years, inflammationand pain in the joints subsided and analgesics and anti-inflammatory drugs could be stopped. The patient never received radiation therapy to the spinal column for AS. A new pain in the L-5, S-1, and S-2 distributions began in 1968, at age 51 years, predominantly on the left side, with severe exacerbations due to mechanical factors. In 1974, water-soluble contrast medium myelography showed an enlarged and lobulated lumbosacral dural sac with no evident radicular compression (Fig 1A). Pain worsened progressively and became chronic and severe. At admission in May 1980, loss of cutaneous sensation on left perianal region, heel, and sole, slight weakness on the left in L-5 innervated muscles, and bilateral ankle areflexia were present. Erythrocyte sedimentation rate was normal. Total protein level was 280 mg/ L, and white cell count was 2 per cubic millimeter in CSF. Lumbosacral mega cul-de-sac was more apparent on metrizamide myelography as well as erosion of posterior walls of the vertebral bodies (scalloping) and a posterior lobulated meningocele predominantly at the L-4 level (Fig 1B). Computed tomographic (CT) scan confirmed erosion of the laminae and posterior spinous processes on their axial side in the lumbosacral region (Fig 2). Electromyography with needle electrode examination showed fibrillations and decreased numbers of motor unit potentials bilaterally in the L-4, L-5, and S-1 root distribution, consistent with neurogenic abnormalities. F waves were delayed or abolished. Nerve motor and sensory conductions were slowed in the lower limbs (30-36 misec). Noramidopyrine had a partial analgesic effect, but transcutaneous electrical stimulation, carbamazepine, steroids, isaxo-
Copyright 0 1991 by the American Neurological Association
221
Fig 2 Lumbar computed tomographic scan at the L-4 level. Enlargement of the vertebral canal and defects in the laminae.
Fig 1. Water-soluble contrast medium myelography. (A) 1974. Enlarged and lobulated caudal sac. (B) 1980. The caudal sac seems larger with posterior meningocele and vertebra/ scalloping.
nine, and B vitamins were without benefit. In 1983, lightning pains and chronic crushing pain predominantly in the left perianal region, heel, and sole became intense. They were worsened by ambulation, which was restricted to 500 m with a cane. Clinical examination was unchanged by comparison with 1780. Radionuclide cisternography was performed with 2 mCi Indium 111 injected at the C-2 level. The cerebral cisternae and lumbosacral dural sac were visualized within normal delays (Fig 3A), but isotope was still detectable in the dural sac at 24 hours. Half-life elimination time of the isotope at this level was markedly increased (1 1 hours; normal range, 1-4 hours). Acetazolamide, up to 500 mg per day, reduced pain intensity for only a 6-month period. A lumboperitoneal shunt was placed in April 1984. Subsequent improvement was pronounced with disappearance of lightning pains, reduced intensity of the chronic pain, and ability to walk up to 2 km. Neurological examination and dural mega cul-de-sac on C T scan, however, were unchanged. Radionuclide cisternography was repeated in Qctober 1984. Isotope rapidly diffused toward lumbosacral dural sac, peritoneal cavity, and bladder (Fig 3B). No intrathecal isotope was detectable at 24 hours after injection. Isotope half-time elimination was 1 hour. During the following years, symptoms fluctuated with periods of intense pain in the legs and inability to walk without external assistance alternating with periods of complete symptomatic remission. Radionuclide cisternography was checked in May 1986 and confirmed that the lumboperito-
222 Annals of Neurology
Vol 27 No 2 February 1771
Fig 3. Indium 111 radionuclide cistwnography. Isotope injection at the C-2 level. (A) 1983. Prolonged stagnation of the isotope in the caudal sac at 1 I and 24 hours after isotope injection. (B) October 1984, 6 months after lumboperitoneal shunt insertion; rapid dz$fusion of the isotope from the caudal sac to the bladder through the bmboperitoneal shunt (L.P.D.).
neal shunt was working properly. Since January 1788, pain has been moderate and ambulation full.
Discussion We propose that a hydrodynamic component may be considered in CES-AS syndrome. Based on the two myelographies performed at a 6-year interval, the dural sac increased in size. Acetazolamide benefited the neurological symptoms although only temporarily. Cisternography demonstrated stagnation of isotope inside the d u a l sac. Lumboperitoneal shunt was able to correct cisternographic abnormalities and, to a lesser extent, neurological symptoms. CES signs were unmodified, but they had been long-standing and were probably irreversible. Defective resorption of lumbar CSF, which physiologically takes place through arachnoid villi in epidural veins 1131, could contribute to dural sac enlargement.
Furthermore, rapid CSF pressure changes that were not efficiently damped [14] insofar as the epidural space was atrophic and the epidural veins were compressed could have contributed to nerve root injury. According to these data, CES-AS is in some respects a normal pressure-enlarged dural sac syndrome, especially as opening lumbar CSF pressure has always been found normal [I, 4 ) in this clinical setting. CES-AS bony erosion concerns only the inner surface and preserves the outer surface of the vertebral canal. This feature could be relevant to the hydrodynamic theory, and “erosion” may not be the appropriate term. Owing to the enlargement of the dural sac and the relative CSF hypertension, a slightly centrifugal displacement of bone formation during continuous bone turnover could be postulated. This dynamic theory needs confirmation. Continuous recording of lumbar CSF pressure in patients with CES-AS and cisternographic studies performed at the inflammatory stage of AS before CES-AS onset would be relevant. In any event, this theory provides a new approach to CES-AS therapy; early CSF shunting may be considered for CES-AS prevention.
3.
4.
5. 6.
7.
8.
9. 10.
11.
We thank Dr G. Veillas, MD, Centre de Midecine Nucleaire, Hbpital Neurologique de Lyon, for technical collaboration.
12.
References
13.
1. Babinski J. Pseudo tabes spondylosique. Rev Neurol 1901; 11645-646 2. Brun C. Le syndrome de la queue de cheval de la spondylarthrite ankylosante (i propos de deux observations person-
14.
nelles). ThSse medecine, UniversitC Claude Bernard, Lyon, 1989, no. 275. Auquier C, Siaud JR, Guiot G, Lhermitte F. Syndrome de la queue de cheval au cours de la spondylarthrite ankylosante: dkcouverte opiratoire d’une fusion entre vertebre et dure-mere dans le cas rapporti. Rev Rhum 1974;41:733-737 Hauge T. Chronic rheumatoid polyarthritis and spondylarchritis associated with neurological symptoms and signs occasionally simulating an intraspinal expansive process. Acta Chir Scand 1961;120:395-401 Lee MLH, Waters DJ. Neurological complication of ankylosing spondylitis. Br Med J 1962;1:798 Matthews WB. The neurological complications of ankylosing spondylitis. J Neurol Sci 1968;6:561-573 Bartleson JD, Cohen MD, Harrington TM, et al. Cauda equina syndrome secondary to long-standing ankylosing spondylitis. Ann Neurol 1983;14:662-669 Pouget J, Guerra JL, Acquaviva PC, Serratrice G. Syndrome de la queue de cheval avec mal perforant plantaire et ostkoarthropathies neurogbes au cours d’une pelvispondylite rhumatismale. Rev Rhum 1985;52:451-455 Gordon AL, Yudell A. Cauda equina lesion associated with rheumatoid spondylitis. Ann Intern Med 1973;78:555-557 Martin G, Luken MD, Dushyant V, et al. Sympcomatic spinal stenosis associated with ankylosing spondylitis. Neurosurgery 1982;11:703-705 Soeur M, Monsen G, Baleriaux-Waha D, et al. Cauda equina syndrome in ankylosing spondylitis. Anatomical, diagnostic and therapeutic considerations. Acta Neurochir (Wien) 1981;55: 303-315 Byrne JV. Cauda equina syndrome in ankylosing spondylitis. Semin Roentgenol 1086;21:lOl-102 Fishman RA. Cerebrospinal fluid in diseases ofthe nervous system. Philadelphia: WB Saunders, 1980:15-16 Gilland 0, Chin F, Anderson WB, Nelson JR. A cinemyelographic study of cerebrospinal fluid dynamics. Am J Roentgenol 1969;1O6:369-3 7 5
Brief Communication: Confavreux et al: CSF Hydrodynamics in CES
223
11.
12.
13.
14.
15.
rents. Magn Resun Med 1990;14:26-30 Frahm J, Bruhn H, Gyngell ML, et al. Localized proton NMR spectroscopy in different regions of the human brain in vivo. Relaxation times and concentrations of cerebral metabolites. Magn Reson Med 1989;11:47-63 Petroff OA, Spencer DD, Alger JR, Pritchard JW. High-field proton magnetic resonance spectroscopy of human cerebrum obtained during surgery for epilepsy. Neurology 1989;39: 1197-2202 Medina L, Chi TL, DeVivo DC, Hilal SK. MR findings in patients with subacute necrotizing encephalomyelopathy (Leigh syndrome): Correlation with biochemical deficit. AJNR 1990; 11:379-384 Frahm J, Michaelis T, Merboldt K-D, et al. Localized NMR spectroscopy in vivo. NMR Biomed 1989;2:188-195 Birken DL, Oldendorf WH. N-acetyl-L-aspartic acid: a literature review of a compound prominent in ‘H-NMR spectroscopic studies of brain. Neurosci Biobehav Rev 1989;13:23-31 Tofts PS, Wray S. A critical asessment of methods of measuring metabolite concentrations by NMR spectroscopy. NMR Biomed 1988;l:l-10
Cerebrospinal Fluid Dynamics in the Tardive Cauda Equina Syndrome of Ankylosing Spondylitis Christian Confavreux, MD,” Jean-Paul Iarbre, MD, 1 Edouard Lejeune, MD,t Marc Sindou, MD,$ and Gilbert Aimard*
Typical cauda equina syndrome secondary to longstanding ankylosing spondylitis is reported in a 63-yearold man. Radionuclide cisternography demonstrated a resorption defect of cerebrospinal fluid in the enlarged lumbosacral dural sac. After transient symptomatic improvement with acetazolamide, a lumboperitoneal shunt was placed. T h e rate of cerebrospinal fluid, isotope resorption became normal. In the 5 years of follow-up, partial remission has been observed. Confavreux C, Larbre J-P, Lejeune E, Sindou M, Aimard G. Cerebrospinal fluid dynamics in the tardive cauda equina syndrome of ankylosing spondylitis. Ann Neurol 1991;29:221-223
Tardive cauda equina syndrome in ankylosing spondylitis (CES-AS) is a rare but well-known entity since From the Tlinique de Neurologie, HBpital Neurologique, and $Service de Neurochirurgie, HBpital Neurologique, Lyon, and +Service de Rhumatologie, HBpitd des Charpennes, Villeurbanne, France. Received Apr 10, 1990, and in revised form Jun 20 and Aug 8, 1990. Accepted for publication Aug 10, 1990. Address correspondence to Dr Confavreux, Clinique de Neurologie, HBpital Neurologique, 59 Boulevard Pinel, 69003 Lyon, France.
Babinski’s pioneer work [l}. To date, 69 cases have been reported 12-7}. Inflammation and fibrosis predominating o n dura matter and epidural space associated with enlargement of the caudal sac, narrowing of epidural space, adherence of dura matter to adjacent bony and ligamentous structures, and bony erosions of laminae and spinous processes are the basic radiological [Z, 81, surgical 13-5, 7, 9-11], and pathological C6, 121 features. The long interval separating the onset of ankylosing spondylitis (AS) and cauda equina syndrome (CES) and the common finding that AS has been burned out for several years when CES begins suggests that active inflammation is not directly involved in its pathogenesis. Alternatively, retractile fibrosis in the epidural space could lead t o a “between-trunk-andbark” variety of entrapment syndrome. We report a patient, with study of the cerebrospinal fluid (CSF) dynamics by isotopic techniques. A resorption defect of the isotope from the caudal sac was observed. Correction by lumboperitoneal shunting resulted in partial improvement of radicular pain.
Patient Report A 63-year-old, HLA-B27 positive, graduate engineer suffered from a relapsing form of AS since the age of 23 years. Sacroiliac joints and spine became fused with pronounced kyphosis. In 1956, vertebral osteotomy with posterior osseous graft was performed at the T12-L1 level for correction of the spinal deformation. During the subsequent years, inflammationand pain in the joints subsided and analgesics and anti-inflammatory drugs could be stopped. The patient never received radiation therapy to the spinal column for AS. A new pain in the L-5, S-1, and S-2 distributions began in 1968, at age 51 years, predominantly on the left side, with severe exacerbations due to mechanical factors. In 1974, water-soluble contrast medium myelography showed an enlarged and lobulated lumbosacral dural sac with no evident radicular compression (Fig 1A). Pain worsened progressively and became chronic and severe. At admission in May 1980, loss of cutaneous sensation on left perianal region, heel, and sole, slight weakness on the left in L-5 innervated muscles, and bilateral ankle areflexia were present. Erythrocyte sedimentation rate was normal. Total protein level was 280 mg/ L, and white cell count was 2 per cubic millimeter in CSF. Lumbosacral mega cul-de-sac was more apparent on metrizamide myelography as well as erosion of posterior walls of the vertebral bodies (scalloping) and a posterior lobulated meningocele predominantly at the L-4 level (Fig 1B). Computed tomographic (CT) scan confirmed erosion of the laminae and posterior spinous processes on their axial side in the lumbosacral region (Fig 2). Electromyography with needle electrode examination showed fibrillations and decreased numbers of motor unit potentials bilaterally in the L-4, L-5, and S-1 root distribution, consistent with neurogenic abnormalities. F waves were delayed or abolished. Nerve motor and sensory conductions were slowed in the lower limbs (30-36 misec). Noramidopyrine had a partial analgesic effect, but transcutaneous electrical stimulation, carbamazepine, steroids, isaxo-
Copyright 0 1991 by the American Neurological Association
221
Fig 2 Lumbar computed tomographic scan at the L-4 level. Enlargement of the vertebral canal and defects in the laminae.
Fig 1. Water-soluble contrast medium myelography. (A) 1974. Enlarged and lobulated caudal sac. (B) 1980. The caudal sac seems larger with posterior meningocele and vertebra/ scalloping.
nine, and B vitamins were without benefit. In 1983, lightning pains and chronic crushing pain predominantly in the left perianal region, heel, and sole became intense. They were worsened by ambulation, which was restricted to 500 m with a cane. Clinical examination was unchanged by comparison with 1780. Radionuclide cisternography was performed with 2 mCi Indium 111 injected at the C-2 level. The cerebral cisternae and lumbosacral dural sac were visualized within normal delays (Fig 3A), but isotope was still detectable in the dural sac at 24 hours. Half-life elimination time of the isotope at this level was markedly increased (1 1 hours; normal range, 1-4 hours). Acetazolamide, up to 500 mg per day, reduced pain intensity for only a 6-month period. A lumboperitoneal shunt was placed in April 1984. Subsequent improvement was pronounced with disappearance of lightning pains, reduced intensity of the chronic pain, and ability to walk up to 2 km. Neurological examination and dural mega cul-de-sac on C T scan, however, were unchanged. Radionuclide cisternography was repeated in Qctober 1984. Isotope rapidly diffused toward lumbosacral dural sac, peritoneal cavity, and bladder (Fig 3B). No intrathecal isotope was detectable at 24 hours after injection. Isotope half-time elimination was 1 hour. During the following years, symptoms fluctuated with periods of intense pain in the legs and inability to walk without external assistance alternating with periods of complete symptomatic remission. Radionuclide cisternography was checked in May 1986 and confirmed that the lumboperito-
222 Annals of Neurology
Vol 27 No 2 February 1771
Fig 3. Indium 111 radionuclide cistwnography. Isotope injection at the C-2 level. (A) 1983. Prolonged stagnation of the isotope in the caudal sac at 1 I and 24 hours after isotope injection. (B) October 1984, 6 months after lumboperitoneal shunt insertion; rapid dz$fusion of the isotope from the caudal sac to the bladder through the bmboperitoneal shunt (L.P.D.).
neal shunt was working properly. Since January 1788, pain has been moderate and ambulation full.
Discussion We propose that a hydrodynamic component may be considered in CES-AS syndrome. Based on the two myelographies performed at a 6-year interval, the dural sac increased in size. Acetazolamide benefited the neurological symptoms although only temporarily. Cisternography demonstrated stagnation of isotope inside the d u a l sac. Lumboperitoneal shunt was able to correct cisternographic abnormalities and, to a lesser extent, neurological symptoms. CES signs were unmodified, but they had been long-standing and were probably irreversible. Defective resorption of lumbar CSF, which physiologically takes place through arachnoid villi in epidural veins 1131, could contribute to dural sac enlargement.
Furthermore, rapid CSF pressure changes that were not efficiently damped [14] insofar as the epidural space was atrophic and the epidural veins were compressed could have contributed to nerve root injury. According to these data, CES-AS is in some respects a normal pressure-enlarged dural sac syndrome, especially as opening lumbar CSF pressure has always been found normal [I, 4 ) in this clinical setting. CES-AS bony erosion concerns only the inner surface and preserves the outer surface of the vertebral canal. This feature could be relevant to the hydrodynamic theory, and “erosion” may not be the appropriate term. Owing to the enlargement of the dural sac and the relative CSF hypertension, a slightly centrifugal displacement of bone formation during continuous bone turnover could be postulated. This dynamic theory needs confirmation. Continuous recording of lumbar CSF pressure in patients with CES-AS and cisternographic studies performed at the inflammatory stage of AS before CES-AS onset would be relevant. In any event, this theory provides a new approach to CES-AS therapy; early CSF shunting may be considered for CES-AS prevention.
3.
4.
5. 6.
7.
8.
9. 10.
11.
We thank Dr G. Veillas, MD, Centre de Midecine Nucleaire, Hbpital Neurologique de Lyon, for technical collaboration.
12.
References
13.
1. Babinski J. Pseudo tabes spondylosique. Rev Neurol 1901; 11645-646 2. Brun C. Le syndrome de la queue de cheval de la spondylarthrite ankylosante (i propos de deux observations person-
14.
nelles). ThSse medecine, UniversitC Claude Bernard, Lyon, 1989, no. 275. Auquier C, Siaud JR, Guiot G, Lhermitte F. Syndrome de la queue de cheval au cours de la spondylarthrite ankylosante: dkcouverte opiratoire d’une fusion entre vertebre et dure-mere dans le cas rapporti. Rev Rhum 1974;41:733-737 Hauge T. Chronic rheumatoid polyarthritis and spondylarchritis associated with neurological symptoms and signs occasionally simulating an intraspinal expansive process. Acta Chir Scand 1961;120:395-401 Lee MLH, Waters DJ. Neurological complication of ankylosing spondylitis. Br Med J 1962;1:798 Matthews WB. The neurological complications of ankylosing spondylitis. J Neurol Sci 1968;6:561-573 Bartleson JD, Cohen MD, Harrington TM, et al. Cauda equina syndrome secondary to long-standing ankylosing spondylitis. Ann Neurol 1983;14:662-669 Pouget J, Guerra JL, Acquaviva PC, Serratrice G. Syndrome de la queue de cheval avec mal perforant plantaire et ostkoarthropathies neurogbes au cours d’une pelvispondylite rhumatismale. Rev Rhum 1985;52:451-455 Gordon AL, Yudell A. Cauda equina lesion associated with rheumatoid spondylitis. Ann Intern Med 1973;78:555-557 Martin G, Luken MD, Dushyant V, et al. Sympcomatic spinal stenosis associated with ankylosing spondylitis. Neurosurgery 1982;11:703-705 Soeur M, Monsen G, Baleriaux-Waha D, et al. Cauda equina syndrome in ankylosing spondylitis. Anatomical, diagnostic and therapeutic considerations. Acta Neurochir (Wien) 1981;55: 303-315 Byrne JV. Cauda equina syndrome in ankylosing spondylitis. Semin Roentgenol 1086;21:lOl-102 Fishman RA. Cerebrospinal fluid in diseases ofthe nervous system. Philadelphia: WB Saunders, 1980:15-16 Gilland 0, Chin F, Anderson WB, Nelson JR. A cinemyelographic study of cerebrospinal fluid dynamics. Am J Roentgenol 1969;1O6:369-3 7 5
Brief Communication: Confavreux et al: CSF Hydrodynamics in CES
223
