Surgical treatment of syringomyelia. Favorable results with syringoperitoneal shunting

Transcription

1 J Neurosurg 61: , 1984 Surgical treatment of syringomyelia Favorable results with syringoperitoneal shunting NICHOLAS M. BARBARO, M.D., CHARLES B. WILSON, M.D., PHILIP H. GUTIN, M.D., AND MICHAEL S. B. EDWARDS, M.D. Department of Neurological Surgery, School of Medicine, University of California, San Francisco, California o- The authors reviewed the clinical findings, radiological evaluation, and operative therapy of 39 patients with syringomyelia. Syringoperitoneal (SP) shunting was used in 15 patients and other procedures were used in 24 patients. Follow-up periods ranged from l 89 to 12 years. During the period of this study, metrizamide myelography in conjunction with early and delayed computerized tomography scanning replaced all other diagnostic procedures in patients with syringomyelia. Preoperative accuracy for the two procedures was 87%. The most common symptoms were weakness (79%), sensory loss (67%), pain (38%), and leg stiffness (28%). Surgery was most effective in stabilizing or alleviating pain (100%), sensory loss (81%), and weakness (74%); spasticity, headache, and bowel or bladder dysfunction were less likely to be reversed. Approximately 80% of patients with idiopathic and posttraumatic syringomyelia and 70% of those with arachnoiditis improved or stabilized. Better results were obtained in patients with less severe neurological deficits, suggesting the need for early operative intervention. A higher percentage of patients had neurological improvement with SP shunting than with any other procedure, especially when SP shunting was the first operation performed. Patients treated with SP shunts also had the highest complication rate, most often shunt malfunction. These results indicate that SP shunting is effective in reversing or arresting neurological deterioration in patients with syringomyelia. KEY WORDS 9 Chiari malformation 9 spinal cord 9 syringomyelia 9 syrinx 9 shunt T HE natural history of syringomyelia is highly variable. Some patients improve or stabilize without surgery, while others deteriorate even with the most aggressive intervention. 2'3a8 Several different operations have been advocated for the treatment of syringomyelia, including posterior fossa decompression with 2'4'9'1~ or without 8,9,19,23,28,29 closure of a patent central canal, laminectomy with syringostomy, syrinx to subarachnoid (SSA) shunting, 25'28'3~ terminal ventriculostomy, H ventriculoatrial shunting, 16 and syringoperitoneal (SP) shunting. 24 Even though satisfactory results have been reported for most of these procedures, some authors have been critical of these results, a2'34 We reviewed the clinical findings, radiological evaluation, and operative therapy of 39 patients with syringomyelia treated at the University of California, San Francisco, from 1970 to The radiological evaluation and operative therapy evolved considerably during that period. A new SP shunt was used in 15 patients and has become the primary treatment for syringomye- lia at this institution. In this report, we compare the results of SP shunting with the results of other operative procedures for syringomyelia. Clinical Material and Methods Patient Population During the 12-year period of this study, 43 patients with syringomyelia were evaluated. Three patients were not treated at this institution, and one patient died in a motor-vehicle accident 4 months after treatment. Patients with cystic spinal tumors were excluded. The remaining 39 patients are included in this analysis. There were 18 males and 21 females, aged 3 to 60 years (mean 34 years, Fig. 1). Symptoms and Signs The presenting symptoms are summarized in Table 1. Weakness and sensory loss were the most common symptoms, and 15 patients had axial or radicular pain J. Neurosurg. / Volume 61/September,

2 N. M. Barbaro, et al. i- z ill I--.< Q.. ii 0 i11 z FIG Range 3-60 years ~/f/~ Males 18 ~J Females 21 [] NIl, AGE (years) 1. Age and sex distribution of 39 patients with syringomyelia. TABLE 1 Presenting symptoms in 39 patients with syringomyelia Symptoms No. of Cases weakness 31 sensory loss 26 pain 15 leg stiffness 11 headache 2 bowel or bladder dysfunction 4 diplbpia 2 oscillopsia 2 or both. Eleven patients had leg stiffness, two had headaches, and four had bowel or bladder dysfunction. The physical findings are described in Table 2. Classic syringomyelia (dissociated sensory loss in a cape-like distribution, hypotonic arm weakness, and spastic leg weakness) was the most common syndrome, but was found in only 13 of 39 patients. When present, these findings were usually asymmetrical. Unilateral sensory and motor deficits were present in 19 patients; six had purely motor findings, five had sensory disturbances only, and eight had both sensory and motor loss. One patient complained of radicular pain, but had a normal neurological examination. Five patients had trigeminal sensory loss. Etiology The cause of syringomyelia in these patients is summarized in Table 3. Of the 25 patients with "idiopathic" syringomyelia, only 15 were proven by radiological evaluation or operation to have tonsillar ectopia. However, many of these patients were evaluated before metrizamide-enhanced computerized tomography (CT) became available. If signs or symptoms of tonsillar ectopia were absent, the foramen magnum was not FIG. 2. Left: Sagittal reconstruction of postmetrizamide computerized tomography (CT) scan showing tonsillar ectopia in a patient with Chiari I malformation. Right: Axial postmetrizamide CT scan showing an eccentrically placed cervical syrinx. 532 J. Neurosurg. / Volume 61/September, 1984

3 Surgery for syringomyelia TABLE 2 Physical findings in 39 patients with syringomyelia Physical Findings No. of Cases classic syringomyelia 13 unilateral motor & sensory deficits 19 focal sensory disturbances 5 focal motor disturbances 3 spastic quadriparesis 4 paraparesis with sensory level 2 normal 1 always fully evaluated. Thus, some or all of the remaining 10 patients may have had tonsillar ectopia. Radiological Evaluation During the early part of this review, patients were evaluated with Pantopaque myelography, air myelography, or both procedures. Two patients had worsening of their neurological deficits after air myelography, from which one patient never fully recovered. Vertebral angiography was performed in five patients. Later, metrizamide myelography in conjunction with early and delayed metrizamide-enhanced CT scanning was used exclusively. Of 23 patients evaluated in this manner, 20 (87 %) had a preoperative diagnosis of syringomyelia, two (9%) had abnormal but nondiagnostic studies, and one (4%) had a normal radiological evaluation. Computerized tomography scanning with metrizamide provided visualization of the foramen magnum (Fig. 2 left), sparing patients from the risk of morbidity associated with air myelography and vertebral angiography. Delayed metrizamide-enhanced CT scans reliably predicted eccentrically located syrinxes (Fig. 2 right) and allowed the operative approach to be directed to the largest portion of the syrinx, where the spinal cord was thinnest. Percutaneous injection of metrizamide into a syrinx was followed by plain radiography and CT scanning in six patients (Fig. 3). This technique revealed the rostrocaudal extent of the cavity, which was often larger than predicted by clinical examination, myelography, or CT scanning; occasionally, this information allowed a lower level of the spinal cord to be approached operatively, thereby avoiding possible damage to cervical structures. Recently, we have used nuclear magnetic resonance imaging to evaluate the spinal cord and posterior fossa in patients with syringomyelia (Fig. 4). FIG. 3. Plain cervical radiograph after percutaneous injection of metrizamide into the syrinx. The syrinx extends from the top of the cervical spinal cord well into the thoracic region. Multiple septations are visible. TABLE 3 Cause of syringomyelia in 39 patients Cause No. of Patients idiopathic 25 tonsiuar ectopia 15 no tonsillar ectopia 10 arachnoiditis 7 posttraumatic 6 hematomyelia 1 FIG. 4. Sagittal nuclear magnetic resonance image of a patient with a cervical spinal cord syrinx. Arrows indicate rostral and caudal limits of the syrinx. J. Neurosurg. / Volume 61/September,

4 N. M. Barbaro, et al. 15 patients, SSA shunting in 19, Gardner procedure 2,~~ in seven, decompressive laminectomy in two, laminectomy and syrinx aspiration in two, Echols procedure ~5 in two, and cord transection for traumatic paraplegia in one. FIG. 5. Syrinx end of a syringoperitoneal (SP) shunt. The flexible T-tube arms have multiple drainage holes and can be cut to the desired length. The shunt is brought out through the spinal cord and dura at a right angle and is secured with the suture tab. A metal step-up connector is used to adapt the SP shunt to standard peritoneal shunt tubing. Operations Forty-eight operations were performed in these 39 patients. Twenty-nine patients underwent a single procedure; eight patients had two operations and one patient had three. Syringoperitoneal shunting was used in TABLE 4 Outcome by symptom in 39 patients with syringomyelia Symptoms Total Outcome Cases Good Fair Poor weakness sensory loss pain leg stiffness headache bowel or bladder dysfunction TABLE 5 Outcome by cause of syringomyelia in 39 patients Cause of Total Outcome Syringomyelia Cases Good Fair Poor idiopathic tonsillar ectopia no tonsillar ectopia arachnoiditis posttraumatic hematomyelia Syringoperitoneal Shunting The SP shunt* has undergone several modifications. The syrinx portion of the most recent version is shown in Fig. 5. The patient is placed in a semi-sitting position with the head in a rigid fixation device. The skin over the spine, neck, chest, and abdomen is draped as for a ventriculoperitoneal shunt. The following factors are considered in choosing the laminectomy site: the widest portion of the syrinx (thinnest portion of spinal cord), the least potential for new neurological deficit, and the most dependent portion of the syrinx. A single-level total or hemilaminectomy is performed. The operating microscope is used, and a small myelotomy is made into the thinnest portion of the spinal cord, usually the dorsal root entry zone. The T-tube arms are cut to the desired length and inserted into the syrinx. The tubing is brought through a separate hole in the dura and fastened with a suture tab to the paraspinal musculature. A step-up connector is used to attach the shunt to a Foltz flat-bottomed flushing reservoirt placed in a subcutaneous pocket, usually in the infraclavicular thoracic region. Standard techniques are used to insert a standard low-pressure peritoneal catheter into the abdominal cavity. Operative Results The operative results were divided into three categories: good, fair, and poor. The outcome was considered good if there was improvement in neurological signs and symptoms, fair if there was stabilization of neurological deficit, and poor if there was neurological deterioration. Overall, the operative results were good in 17 patients (44%), fair in 13 (33%), and poor in nine (23%). Outcome by Symptoms, Etiology, and Neurological Deficit The effect of operative intervention on the presenting symptoms is summarized in Table 4. Pain was the symptom most frequently affected: 11 (73%) of 15 patients had less pain after surgery, and four (27 %) were stabilized. Weakness was reduced or stabilized in 23 (74%) of 31 patients, and sensory loss was reduced or stabilized in 21 (81%) of 26 patients. Spasticity, headache, and bowel or bladder symptoms were the least likely to be reversed. * Syringoperitoneal shunt manufactured by Heyer-Schulte Corp., Goleta, California. t Reservoir manufactured by Heyer-Schulte Corp., Goleta, California. 534 J. Neurosurg. / Volume 61 /September, 1984

5 9 Surgery for syringomyelia A FOCAL DEFICIT B UNILATERAL Motor N=5 Sensory... N=3 / 9...:? ~"...:..,,"...,.L"~"~"~, :..;;..::-.. Improved 5(63%) t Improved 6(55%) \ \ \ \ < 1 month z 1 Worse 0(0%),ear < 1 month >_ 1 year _Stable 3(37%) N=11 - C CLASSICAL SYRINGOMYELIA D SPASTIC QUADRIPARESIS Improved 5(38.5%) i improve d 010%1 N=13-. Stable 5(38.5%). Stable < 1 month > 1,ear Worse 3(23%) I I / ~-'~...~ Worse I W- ~ 3175%~ < 1 month > 1 year FIG. 6. Comparison of preoperative, immediate postoperative, and long-term results according to preoperative neurological deficit. Early postoperative results did not always predict follow-up results at 1 year. The best results were obtained in patients with the mildest preoperative deficits. Outcome by etiology is shown in Table 5. There was stabilization or improvement in approximately 80% of patients in the idiopathic and posttraumatic groups and in approximately 70% of patients in the arachnoiditis group. The patient with hematomyelia had a poor result. Figure 6 shows the follow-up results at 1 month and 1 year or more after surgery according to the preoperative neurological findings. The best results were obtained in patients with a focal deficit, 63% of whom showed improvement, and in those with a unilateral deficit, of whom 55 % showed improvement. The worst results were in the patients with spastic quadriparesis: three of four were worse at 1 year. Outcome by Type of Operation The 48 operations were divided into three groups to facilitate comparison. The first group (no shunt) consisted of 14 procedures in which cyst fluid was not shunted into the subarachnoid space or to an extraspinal site. The results in this group are listed in Table 6. The second group (SSA shunt) included 19 procedures in which a Silastic catheter was used to drain cyst fluid into the subarachnoid space. The third group (SP shunt) consisted of 15 SP shunting procedures. The outcome for all three groups is shown in Table 7. The classification described above was used, except that the result was considered poor if a patient required another operation because of neurological deterioration. Thus, a patient was listed in more than one group TABLE 6 Outcome of 14 procedures not involving shunts Procedure No. Outcome Good Fair Poor Gardner laminectomy Echols aspiration cord transection total J. Neurosurg. / Volume 61 /September,

6 N. M. Barbaro, et al. Procedure* TABLE 7 Outcome by type of operation No. Good Fair Poor No. % No. % No. % no shunt SSA shunt SP shunt * SSA = syrinx to subarachnoid; SP = syringoperitoneal. if more than one operation was used. The three groups did not differ significantly in average age, duration of symptoms, or cause of syringomyelia. The follow-up period was 2 to 13 years in the no-shunt group, 1 to 10 years in the SSA shunt group, and 1 to 3 years in the SP shunt group. A higher percentage of patients experienced neurological improvement with SP shunting than with any other operation. In Table 8, the SP shunt patients are divided into two groups: those in whom an SP shunt was the primary treatment (including two patients who underwent suboccipital craniectomy for symptomatic Chiari I malformations) and those in whom a prior therapy failed. The best results were in the 10 patients who had a primary SP shunt: eight patients improved and two became stabilized. Complications There were 12 complications in 46 operations (Table 9). In the SSA shunt group, one patient had postoperative meningitis and another had severe postoperative pain requiring several pain-related procedures, including deep-brain stimulation. In the SP shunt group there was one urinary tract infection that resulted in sepsis, two cases of temporary neurological deterioration (paraparesis and incontinence), and six shunt-related complications: one infection requiring shunt removal, one low-pressure headache requiring placement of an antisiphon advice, and four shunt obstructions. Three of the shunt obstructions occurred in the first four patients treated; therefore, the SP shunt design was altered to include tubing of larger diameter. In three cases, the obstruction was at the peritoneal end, and only required revision of the peritoneal tubing. Discussion The clinical findings in our patients with syringomyelia are similar to those reported in previous selies. 2"6'8'10'18'21'22'29'30 Schlesinger, et a/., 29 recently reviewed the clinical findings in 60 patients with syringomyelia. Only a few of them had the classic syndrome of syringomyelia, but unilateral deficits, which were occasionally purely motor or sensory, were relatively common, probably reflecting diagnosis at an earlier stage of the disease. These authors ascribed this finding to greater awareness of the clinical entity of syringomyelia and to improved diagnostic capabilities. TABLE 8 Outcome in the syringoperitoneal shunt group Shunting Procedure* Outcome Good Fair Poor primary SP shunt SP shunt alone craniectomy secondary SP shunt previous SSA shunt SSA plus craniectomy previous laminectomy * sp = syringoperitoneal; SSA = syrinx to subarachnoid. The major improvements in neuroradiological techniques made during the period of this study have greatly facilitated the diagnosis of syringomyelia. Previous diagnostic studies have been replaced by metrizamide myelography and CT scanning, especially the 4- to 6- hour delayed scan. The safety and accuracy of these new techniques are well documented Metrizamide CT scans can visualize the foramen magnum to reveal tonsillar ectopia and can demonstrate asymmetries in the syrinx much more effectively than plain myelography. Percutaneous cyst puncture 6'7'18'29'31 has also been used as a diagnostic and therapeutic maneuver in patients with syringomyelia. Injection of water-soluble contrast material into the syrinx can be given safely and easily, and provides information about the exact rostrocaudal extent of the syrinx. Syrinxes examined by this technique often prove to be larger than predicted by other studies. Nuclear magnetic resonance scanning will probably replace these positive-contrast techniques in the diagnosis and follow-up review of patients with syringomyelia. The pathophysiology of syringomyelia is still debated, and numerous reviews of this subject are available. 2'3's,1~ Most authors agree that the syrinx is subject to continuous pulsatile pressure from outside the spinal cord and to sudden rises in pressure associated with coughing, sneezing, and Valsalva maneuvers. Shunting the contents of the syrinx to a low-pressure extraspinal site offers the best chance of maintaining the pressure within the syrinx at a reduced level and dissipating the effects of sudden pressure changes) 2-34 Some authors disagree with this approach. For example, Levy, et al.,18 reviewed the results of syrinx shunting in 127 patients with Chiari malformation and concluded that shunting offered no advantage over other forms of therapy; however, in their analysis, the results of subarachnoid and extraspinal shunting were combined. Laha, et al., 17 described the shunting of a syrinx to the subarachnoid space using a T-shaped tube. Phillips and Kindt 24 reported preliminary results of syringoperitoneal shunting in four patients. The present study is the largest reported experience with extraspinal syrinx shunting and the first in which a T-shaped SP shunt is used. 536 J. Neurosurg. / Volume 61/September, 1984

7 Surgery for syringomyelia TABLE 9 Complications in each treatment group* Complications No Shunt SSA Shunt SP Shunt shunt obstruction new neurological deficit meningitis pain sepsis shunt infection low-pressure headache total complications * SSA = syrinx to subarachnoid; SP = syringoperitoneal; -- = not applicable. In our series, 77% of patients were stabilized or improved after surgery. These results compare favorably with the results of most published series, in which long-term follow-up data show a 75% stabilization or improvement rate3 's'9'lsz~'28"3~ The neurological examination in the early postoperative period is not a reliable indication of the patient's condition at 1 year. Several of our patients who were stable or worse at 1 month had improved by 1 year, and two patients who were improved at 1 month deteriorated during the long-term follow-up period. Tator, et al., s~ correlated improved clinical outcome with shorter duration of preoperative symptoms. In the present series, the severity of the preoperative neurological deficit was inversely related to outcome. These findings argue strongly for early surgical intervention in patients with syringomyelia, despite occasional reports of patients who have a prolonged stable course without such intervention. Comparison of the different operations performed at our institution showed that SP shunting produced the best results. Among patients who received a primary SP shunt, 53% improved and 27% stabilized. In contrast, less than 50% of patients who deteriorated after one type of treatment were stabilized by secondary SP shunting. These results may reflect the shorter followup experience with SP shunting and improved diagnostic capabilities that allowed earlier treatment. Syringoperitoneal shunting was associated with a higher complication rate than the other operations, but this probably reflects the ease with which SP shunt obstruction is diagnosed. Undoubtedly, some of the failures in the SSA shunt group were caused by shunt obstruction, but the difficulty in assessing SSA shunt patency makes this assertion difficult to prove. Therefore, obstruction is generally not considered in the complication rate of SSA shunting. Syringoperitoneal shunting provided consistently good results, eliminated the need to explore the posterior fossa, and allowed shunt patency to be tested on an outpatient basis. Our results indicate that SP shunting is a simple and effective means of reversing or arresting neurological deterioration in patients with syringomyelia. More experience and longer follow-up monitoring are needed before the role of this operation can be fully evaluated. Acknowledgments We thank the Heyer-Schulte Corp. for assisting with syringoperitoneal shunt design, Tania Retivov for typing the manuscript, and Stephen Ordway for editorial assistance. References 1. Aubin ML, Vignaud J, Jardin C, et al: Computed tomography in 75 clinical cases of syringomyelia. AJNR 2: , Barnett HJM, Foster JB, Hudgson P: Syriogomyelia. Major Problems in Neurology, Vol 1. Philadelphia/London: WB Saunders, Bertrand G: Dynamic factors in the evolution of syringomyelia and syringobulbia. Clin Neurosurg 20: , Cahan LD, Bentson JR: Considerations in the diagnosis and treatment of syringomyelia and the Chiari malformation. J Neurosurg 57:24-31, Conway LW: Hydrodynamic studies in syringomyelia. J Neurosurg 27: , Ellertsson AB: Syringomyelia and other cystic spinal cord lesions. Acta Neurol Scand 45: , Ellertsson AB, Greitz T: Myelocystographic and fluorescein studies to demonstrate communication between intrameduuary cysts and the cerebrospinal fluid space. Acta Neurol Scand 45: , Faulhauer K, Loew K: The surgical treatment of syringomyelia. Long-term results. Acta Neorochir 44: , Garcia-Uria J, Leunda, G, Carrillo R, et al: Syringomyelia: long-term results after posterior fossa decompression. J Neurosurg 54: , Gardner WJ, Angel J: The mechanism of syringomyelia and its surgical correction. Clin Neurosurg 6: , Gardner WJ, Bell HS, Poolos PN, et al: Terminal ventriculostomy for syringomyelia. J Neurosurg 46: , Gardner WJ, McMurry FG: "Non-communicating" syringomyelia: a non-existent entity. Surg Neurol 6: , Hall PV, Muller J, Campbell RL: Experimental hydrosyringomyelia, ischemic myelopathy, and syringomyelia. J Neurosurg 43: , Hall PV, Turner M, Aichinger S, et al: Experimental syringomyelia. The relationship between intraventricular and intrasyrinx pressures. J Neurosurg 52: , Kirgis HD, Echols DH: Syringo-encephalomyelia. Discussion of related syndromes and pathologic processes, with report of a case. J Neurosorg 6: , Krayenbfihl H: Evaluation of the different surgical approaches in the treatment of syringomyelia. Clin Neurol Neurosurg 77: , Laha RK, Malik HG, Langille RA: Post-traumatic syringomyelia. Surg Neurol 4: , Levy WJ, Mason L, Hahn JF: Chiari malformation presenting in adults: a surgical experience in 127 cases. Neurosurgery 12: , Logue V, Edwards MR: Syringomyelia and its surgical treatment -- an analysis of 75 patients. J Neurol Neurosorg Psychiatry 44: , Love JG, Olafson RA: Syringomyelia: a look at surgical therapy. J Neurosurg 24: , McIlroy W J, Richardson JC: Syringomyelia: a clinical review of 75 cases. Can Med Assoc J 93: , 1965 J. Neurosur'g. / Volume 61/September,

8 N. M. Barbaro, et al. 22. Netsky MG: Syringomyelia. A clinicopathologic study. Arch Neuroi Psychiatry 70: , Paul KS, Lye RH, Strang FA, et al: Arnold-Chiari malformation. Review of 71 eases. J Neurosurg 58: , Phillips TW, Kindt GW: Syringoperitoneal shunt for syringomyelia: a preliminary report. Surg Neurol 16: , Pitts FW, Groff RA: Syringomyelia: current status of surgical therapy. Surgery 56: , Pullicino P, Kendall BE: Computed tomography of "cystic" intramedullary lesions. Neuroradioiogy 23: , Resj6 IM, Harwood-Nash DC, Fitz CR, et al: Computed tomographic metrizamide myelography in syringohydromyelia. Radiology 131: , Rhoton AL Jr: Micosurgery of Arnold-Chiari malformation in adults with and without hydromyelia. J Neurosurg 45: , Schlesinger EB, Antunes JL, Michelsen W J, et al: Hydromyelia: clinical presentation and comparison of modalities of treatment. Neurosurgery 9: , Tator CH, Meguro K, Rowed DW: Favorable results with syringosubarachnoid shunts for treatment of syringomyelia. J Neurosnrg 56: , Westberg G: Gas myelography and percutaneous puncture in the diagnosis of spinal cord cysts. Aeta Radiol Suppl 252:1-67, Williams B: A critical appraisal of posterior fossa surgery for communicating syringomyelia. Brain 101: , Williams B: The distending force in the production of "communicating syringomyelia." Lancet 2: , Williams B, Fahy G: A critical appraisal of "terminal ventriculostomy" for the treatment of syringomyelia. J Neurosurg 58: , 1983 Manuscript received February 1, Address reprint requests to: Michael S. B. Edwards, M.D., % The Editorial Office, Department of Neurological Surgery, 350 Parnassus, Suite 807, San Francisco, California J. Neurosurg. / Volume 61/September, 1984