Posttraumatic spinal cord tethering and syringomyelia: surgical treatment and long-term outcome

Transcription

1 J Neurosurg Spine 11: , 2009 Posttraumatic spinal cord tethering and syringomyelia: surgical treatment and long-term outcome Clinical article Sc o t t P. Fa l c i, M.D., 1 Ch a r l o t t e Ind e c k, M.S.N., 1 a n d Da n i e l P. La m m e r t s e, M.D. 2,3 Departments of 1 Neurosurgery and 2 Physical Medicine and Rehabilitation, Craig Hospital, Englewood; and 3 University of Colorado Denver and Health Sciences Center, Aurora, Colorado Object. Permanent neurological loss after spinal cord injury (SCI) is a well-known phenomenon. There has also been a growing recognition and improved understanding of the pathophysiological mechanisms of late progressive neurological loss, which may occur after SCI as a result of posttraumatic spinal cord tethering (SCT), myelomalacia, and syringomyelia. A clinical study of 404 patients sustaining traumatic SCIs and undergoing surgery to arrest a progressive myelopathy caused by SCT, with or without progressive myelomalacia and cystic cavitation (syringomyelia) was undertaken. Both objective and subjective long-term outcomes were evaluated. To the authors knowledge, this is the first series of this size correlating long-term patient perception of outcome with long-term objective outcome analyses. Methods. During the period from January 1993 to November 2003, 404 patients who had previously sustained traumatic SCIs underwent 468 surgeries for progressive myelopathies attributed to tethering of the spinal cord to the surrounding spinal canal, with or without myelomalacia and syrinx formation. Forty-two patients were excluded because of additional pathological entities that were known to contribute to a progressive myelopathy. All surgeries were performed by the same neurosurgeon at a single SCI treatment center and by using a consistent surgical technique of spinal cord detethering, expansion duraplasty, and when indicated, cyst shunting. Results. Outcome data were collected up to 12 years postoperatively. Comparisons of pre- and postoperative American Spinal Injury Association sensory and motor index scores showed no significant change when only a single surgery was required (86% of patients). An outcome questionnaire and phone interview resulted in > 90% of patients self-assessing arrest of functional loss; > 50% of patients self-assessing improvement of function; 17 and 18% selfassessing improvement of motor and sensory functions to a point greater than that achieved at any time postinjury, respectively; 59% reporting improvement of spasticity; and 77% reporting improvement of hyperhidrosis. Conclusions. Surgery for spinal cord detethering, expansion duraplasty, and when indicated, cyst shunting, is a successful treatment strategy for arresting a progressive myelopathy related to posttraumatic SCT and syringomyelia. Results suggest that surgery leads to functional return in ~ 50% of patients, and that in some patients posttraumatic SCT limits maximal recovery of spinal cord function postinjury. A patient s perception of surgery s failure to arrest the progressive myelopathy corresponds closely with the need for repeat surgery because of retethering, cyst reexpansion, and pseudomeningocele formation. (DOI: / SPINE09333) Ke y Wo r d s syringomyelia spinal cord tethering cyst spinal cord injury Pe r m a n e n t neurological loss subsequent to SCI from any cause is a well-known phenomenon. The delayed sequela of progressive cyst/syrinx formation (PPCM) along with progressive neurological loss has long been established. 2 5,7 15,17,18,20 32 More recently, it has been recognized that progression of neurological loss can occur due to scarring or tethering of the spinal cord to the surrounding dura mater without cyst formation Abbreviations used in this paper: ASIA = American Spinal Injury Association; PPCM = progressive posttraumatic cystic myelopathy; PPNM = progressive posttraumatic noncystic myelopathy; SCI = spinal cord injury; SCT = spinal cord tethering. (PPNM). 8,10,12,16,19,23 It has been suggested that posttraumatic SCT is a necessary precursor to cyst formation and progression, and that both entities (PPCM and PPNM) show similar pathophysiological mechanisms, with cyst formation being an end point of a process. 3,4,6,8,10,12,14,23 A retrospective and prospective analysis was performed in patients treated with a consistent surgical technique of spinal cord detethering, expansion duraplasty, and when indicated by specific criteria, cyst shunting. Emphasis was placed on both short- and long-term neurological outcome assessed using ASIA sensory and motor scoring systems 1 as well as patient self-assessment by questionnaire. To the best of our knowledge, this is the first series J Neurosurg: Spine / Volume 11 / October

2 S. P. Falci, C. Indeck, and D. P. Lammertse of this size correlating long-term patient perception of outcome with long-term objective outcome analyses. Methods Study Participants During the period from January 1993 to November 2003, 404 patients who had sustained traumatic SCIs underwent 468 surgeries for progressive myelopathies attributed to tethering of the spinal cord to the surrounding spinal canal, with or without myelomalacia and cyst formation. All surgeries were performed by the same neurosurgeon at 1 inpatient SCI treatment center. Fortytwo patients were deemed to be ineligible for this internal review board approved study because of additional pathological entities, including Chiari I malformation, hydrocephalus, and spinal cord tumor. Therefore, 362 patients were otherwise deemed eligible and were the focus of this investigation. Each of the 362 eligible patients had varying degrees of SCT, myelomalacia, and/or cystic cavitation demonstrated on MR imaging, cine MR imaging, and/or CT myelography. Categories of traumatic events leading to SCI included motor vehicle accidents, aircraft accidents, falls, athletic activities, gunshot wounds, hemorrhage, and previous surgery to the spine and spinal cord. Nearly half (45%) of injuries were cephalad to C-6, 23% were from C-6 to T-1, and 32% were caudal to T-1. Initial neurological classification of the patients was ASIA Impairment Scale Grade A in 63.3%, Grade B in 9.9%, Grade C in 11.3%, Grade D in 14%, and Grade E in 0.6%. The average patient age at the time of surgery was 40.5 years (range years), and the average time of presentation for surgery postinjury was 10.7 years (range 4 weeks 43.8 years). In 23.5% of patients, 1 4 surgeries had been performed in their spinal cords for treatment of a progressive posttraumatic myelopathy at another institution prior to presentation. Symptomatology at presentation included new development and progression in severity of the following: motor loss, sensory loss, functional loss, spasticity, neuropathic pain, autonomic dysreflexia, hyperhidrosis, Horner syndrome, hypothermia, headache, blurred vision, cognitive impairment, extremity edema, and dermatological changes such as piloerection and patchy erythema blushing. A retrospective analysis of outcomes at 1 year postsurgery for 59 of these patients was previously reported. 12 Data Collection Detailed sensory evaluation of all skin dermatomes in light touch and pinprick modalities and detailed motor evaluation of all muscle groups were performed by experienced SCI center occupational and physical therapists independent of the surgical team (Figs. 1 and 2). The ASIA sensory and motor scores were abstracted from these data by researchers independent of the surgical team. The ASIA motor scoring was available in 321 patients (87% of eligible cases), light touch in 337 patients (93% of eligible cases), and pinprick in 339 patients (94% of eligible cases), from 1 to 60 days preoperatively. The ASIA motor, light touch, and pinprick scoring was available in groupings closest to 60 days, 365 days, 3 years, and as far out as 12 years postoperatively, as described in Tables 1 and 2. The ASIA motor, light touch, and pinprick scoring was available in the 313 patients (86% of eligible cases) who underwent a single surgery during the course of the study, as described in Tables 3 and 4. In general, ASIA sensory and motor scoring was not available during the time frame spanning the date of injury to the first preoperative testing, because the majority of patients did not undergo their initial rehabilitation or medical followup at our institution. Additional Outcome Measures Because ASIA scoring does not assess many muscle groups and because it assesses only specific points on a sensory dermatome, the scoring system often does not reflect a significant functional change in the patient (Fig. 3). For these reasons, a patient-centered outcome measure was deemed necessary. A questionnaire was developed to determine the patient s perception of postoperative outcome regarding sensory and motor function, spasticity, hyperhidrosis, pain, or other neurological signs/ symptoms; the patient s perception of whether any lost functions returned following the surgery; and a global assessment of perceived benefit from the surgery. Attempts were made to contact each of the 362 eligible patients by telephone to administer the outcome questionnaire. Following informed consent for participation, responses were obtained from 250 (69%) of the eligible patients, who were as many as 11 years postsurgery. The questionnaire was administered by researchers independent of the surgical team. The ASIA motor, light touch, and pinprick scoring was available in 242 of the 250 patients who responded to the questionnaire as described in Tables 5 and 6. Of these 250 patients, 152 reported progressive loss of strength preoperatively. One hundred twenty-six (83%) of the 152 underwent only a single surgery during the study and had ASIA motor, light touch, and pinprick scoring available as described in Tables 7 and 8. Of the 250 patients responding to the questionnaire, 128 reported progressive loss of sensation preoperatively; 103 (80%) of these 128 underwent only a single surgery during the study, and had ASIA motor, light touch, and pinprick scoring available as described in Tables 9 and 10. Data Analysis Analyses of available data were conducted in 2 groups of individuals: the subgroup of patients who required only 1 surgery by the investigating surgeon during the study, and the group of all eligible patients (who required 1 surgery during the period of the study). Patients who required > 1 surgery were thought to represent a subcategory of individuals who were at higher risk of postoperative retethering, cyst reexpansion, and additional progression of signs and symptoms. In addition, some operations on patients in the multiple procedures group were performed at other centers by other surgeons and thus did not necessarily represent an accurate reflection of consistent application of the investigating surgeon s operative techniques. 446 J Neurosurg: Spine / Volume 11 / October 2009

3 Posttraumatic spinal cord tethering and syringomyelia Fig. 1. A: Detailed record of light touch testing performed in a patient with SCT. B: Detailed record of pinprick testing performed in the same patient. Note that in both panels only the black dots are the points used in ASIA scoring. OT = occupational therapy; PT = physical therapy. Fig. 2. Detailed records of upper-extremity motor function in all muscle groups of a patient with SCT, including hand grasp and pinch evaluation, as well as lower-extremity motor function in the same patient. Note that only the shaded muscle groups in both charts are used in ASIA scoring. J Neurosurg: Spine / Volume 11 / October

4 S. P. Falci, C. Indeck, and D. P. Lammertse All data analyses were performed using SPSS version 13.0 software. Paired-samples t-tests were used to compare pre- and postoperative ASIA sensory and motor scores documented within 1 60 days of the surgery, and again in those with additional documentation, up to 12 years postoperatively. To determine if there were differences in outcomes for the subset of patients who required cyst shunting, differences between pre- and postoperative sensory and motor scores were calculated. Independent-samples t-tests were then used to compare the extent of pre- and postoperative sensory and motor change between patients with and without cyst shunting. For all t-test analyses, a 2-tailed significance level of 0.05 was designated. To determine if there were differences in outcomes related to level of injury, age at surgery, years postinjury, and sex, linear regressions analyses were performed. Subjective reports of long-term postsurgical outcomes obtained via questionnaire up to 11 years following surgery were analyzed using descriptive techniques. Neuroimaging Studies Preoperative imaging studies included plain radiographs and CT scans obtained to assess spinal stability and MR imaging performed to assess the degree of SCT, myelomalacia, and cystic cavitation. Spinal cord tethering was recognized in MR imaging as absence of subarachnoid space or abnormal shape and position of the spinal cord within the spinal canal (Figs. 4 6). Myelomalacia was recognized as a widened spinal cord with low signal in the T1-weighted images, signal hyperintensity to CSF on proton-density images, and increased signal on T2- weighted images. Cystic cavitation or syrinx formation was recognized as a signal that was isointense with CSF on T1-weighted, proton-density, and T2-weighted images. If metal artifact did not allow adequate assessment by MR imaging of the spinal canal and spinal cord, then delayed CT myelography was performed. With CT myelography, SCT was recognized as an absence of subarachnoid space or abnormal positioning of the spinal cord within the spinal canal, 12 and cystic cavitation was documented by evidence of myelographic dye within the cord. All patients were evaluated at surgery by using intraoperative ultrasonography to assess for cyst size, configuration, and location; regions of myelomalacia; location of SCT; and motion of the cord and rootlets. Tethering was recognized as a loss of normal subarachnoid space, close approximation of the cord and rootlets to the dura mater, and the loss of the normal anterior posterior motion of the spinal cord synchronous with the patient s heartbeat and CSF pulsation. Myelomalacia was recognized as regions of hyperechoicity, although severe myelomalacia appeared as regions of hypoechoicity. A cyst cavity or syrinx was recognized as an area of anechoicity. 10,12 TABLE 1: The ASIA motor and sensory index scores in all 362 eligible patients with posttraumatic SCT and syringomyelia* ASIA Score Testing Category No. of Pts Mean Median SD Min Max preop days from op motor strength light touch pinprick st FU days from op motor strength light touch pinprick yr FU days from op motor strength light touch pinprick last available FU days from op motor strength light touch pinprick * FU = follow-up; pts = patients. Preoperative testing was done between 0 and 60 days before surgery; closest to 0 days if multiple tests were performed. First postoperative follow-up testing was done between 10 and 60 days after surgery; closest to 60 days if multiple tests were performed. The 1-year follow-up testing was done between 180 and 545 days after surgery; closest to 365 days if multiple tests were performed. The last follow-up test denotes the latest available test done 180 days postsurgery. Surgical Methods The surgical objectives included the following: 1) decrease of traction of the cord and rootlets by scar; 2) improvement in spinal cord and rootlet mobility demonstrated by increased motion with CSF pulsation; 3) improved CSF flow; and 4) in those patients with cyst cavities, facilitation of the egress of cyst fluid. Laminectomies were performed to expose the dura mater in regions of SCT as determined by preoperative imaging studies. Intraoperative ultrasonography was then performed to ensure adequate exposure of normal subarachnoid space cephalad and caudal to regions of SCT, as well as to assess regions of myelomalacia and cystic cavitation. The dura mater was then opened and the operative microscope was used to aid in dissection of arachnoidal scar that was tethering the spinal cord and rootlets to the surrounding dura. Spinal cord detethering was performed dorsally, laterally, anterolaterally, and widely to the level of the nerve root foramen. Attention was given to the reestablishment of cord and rootlet motion with CSF pulsation, and release of tension off these elements. Anterior and posterior subarachnoid spaces were entered cephalad, caudal, and along regions of tethering to improve CSF flow in both craniocaudal and caudocranial directions. If SCT existed anteriorly, detethering was performed to the limits of a posterior exposure. If a spinal cord cyst was present and did not collapse subsequent to cord detether- 448 J Neurosurg: Spine / Volume 11 / October 2009

5 Posttraumatic spinal cord tethering and syringomyelia TABLE 2: The pre- to postoperative changes in ASIA motor and sensory index scores in all 362 eligible patients with posttraumatic SCT and syringomyelia ASIA Score (mean ± SD) Testing Category* No. of Pts Pretest Posttest Difference p Value preop to 1st postop test motor strength ± ± ± light touch ± ± ± pinprick ± ± ± preop to 1 yr motor strength ± ± ± light touch ± ± ± pinprick ± ± ± preop to last FU motor strength ± ± ± light touch ± ± ± pinprick ± ± ± ing as evaluated by ultrasonography, and certainly if the cyst expanded with Valsalva maneuvers, then a myelotomy was performed in the midline, at the most caudal end of the cyst, for placement of a shunt tube. In general, a shunt tube with the caliber of a lumbar drain was used for shunting, but smaller-caliber tubes were used, in some cases even < 1 mm in diameter. Shunt tubes were threaded the entire length of the cyst. These tubes had multiple perforations along their lengths. The distal end of the shunt tube was placed in normal subarachnoid space caudal to the regions of SCT, or in the peritoneal space for larger cysts when it was believed that the craniospinal axis would be unable to reabsorb the large amount of cyst fluid. An expansion duraplasty was subsequently placed to minimize the chance of retethering. Cadaveric dura mater was predominantly used for the duraplasty, but when this was not available, bovine pericardium was used, and rarely, cadaveric fascia lata. The duraplasty was fashioned to recreate near-normal canal dimensions and was tented laterally and at cephalad and caudal ends to minimize the chance of collapse by postoperative epidural blood, seroma, and CSF. Postoperatively, patients were kept on flat bed rest for 5 days to minimize the chance of CSF leakage, routinely turned side to side to minimize the chance of skin breakdown, pulmonary complications, and rescarring, and then gradually mobilized. Electrical Monitoring All surgeries were performed with somatosensory evoked potential monitoring if a baseline signal could be obtained. The averaging of electrical data was greatly shortened so that feedback could be given to the surgeon every seconds while surgical detethering was being performed. Detethering was stopped if amplitude was reduced or latency was increased, and surgery was only continued when signals returned to baseline. J Neurosurg: Spine / Volume 11 / October 2009 Results Objective Outcome Measures: ASIA Sensory and Motor Index Scores A comparison of the ASIA sensory and motor index scores obtained preoperatively in all eligible patients, regardless of presenting symptoms, to scores obtained from 10 to 60 days postoperatively, showed significant improvements (1.4 points on average for light touch scores [p = 0.03] and 1.4 points on average for pinprick scores [p = 0.03]). A nonstatistically significant decrease of 0.30 points on average (p = 0.06) was identified in motor scoring (Tables 1 and 2). A comparison of ASIA sensory and motor index scores obtained on average at 1 year postoperatively, regardless of number of surgeries per patient, failed to show statistically significant changes in light touch, pinprick, or motor scoring. A comparison of ASIA sensory and motor index scores obtained on average at 3.3 years postoperatively, regardless of number of surgeries performed per patient, failed to show statistically significant changes in light touch or pinprick scores, although a statistically significant decrease of 1.2 points in motor scoring (p = 0.02) was found. A comparison of ASIA sensory and motor index scores obtained preoperatively in the group of patients who had only a single surgery during the study, to scores obtained from 10 to 60 days postoperatively, regardless of presenting symptoms, showed significant improvement (1.6 points [p = 0.02] in light touch scores and 1.4 points [p = 0.04] for pinprick scores), although there was no significant change in motor scores (Tables 3 and 4). When ASIA sensory and motor scoring was evaluated for the same group at a mean follow-up of 1 and 3.3 years postoperatively, no statistical difference in light touch, pinprick, or motor scores was found. In the 250 patients who completed outcome questionnaires, a comparison of preoperative ASIA sensory 449

6 S. P. Falci, C. Indeck, and D. P. Lammertse TABLE 3: The ASIA motor and sensory index scores in 313 single-surgery patients with posttraumatic SCT and syringomyelia ASIA Score Testing Category* No. of Pts Mean Median SD Min Max preop days from op motor strength light touch pinprick st FU days from op motor strength light touch pinprick yr FU days from op motor strength light touch pinprick last FU days from op motor strength light touch pinprick and motor index scores to those obtained in the 10- to 60-day postoperative period, regardless of presenting symptoms, showed significant mean improvements (1.6 points [p = 0.01] for light touch scores and 1.9 points [p = 0.01] for pinprick scores). No significant pre- to postoperative change was found for motor scores in this group (Tables 5 and 6). A comparison of pre- and postoperative ASIA sensory and motor index scores obtained at a mean follow-up of 1 year for the same group, regardless of the number of surgeries performed, failed to show significant differences in light touch, pinprick, or motor scores. A comparison of preoperative and postoperative ASIA sensory and motor scores obtained at a mean follow-up of 3.4 years, regardless of the number of surgeries performed, failed to show significant differences in light touch or pinprick scores, although a decrease in motor score of 1.3 points (p = 0.03) was found. Motor and sensory outcomes were also analyzed according to presenting symptoms; 152 of the 250 patients completing the outcome questionnaire reported progressive loss of motor function as a reason they sought neurosurgical treatment. Of these patients, 128 (84%) required only a single surgery during the study period. A comparison of this group s pre- and postoperative ASIA sensory and motor index scores obtained days after surgery showed significant improvements (2.0 points [p = 0.04] for light touch and 2.4 points [p = 0.03] for pinprick), but no difference for motor outcome (Tables 7 and 8). A comparison of pre- and postoperative ASIA sensory and motor index scores obtained at mean follow-up times of 1 and 3.4 years for this group presenting with symptomatic motor loss showed no significant changes. Of the 250 patients completing the outcome questionnaire, 128 reported progressive loss of sensation as a reason they sought neurosurgical treatment. Of these patients, 105 (82%) required only 1 surgery during the study period. A comparison of ASIA sensory and motor scores obtained preoperatively in this group, to scores obtained between 10 and 60 days postoperatively, showed light touch improvement of 2.2 points, which approached significance (p = 0.057), although there were no significant changes in pinprick or motor scores (Tables 9 and 10). A TABLE 4: The pre- to postoperative changes in ASIA motor and sensory index scores in 313 single-surgery patients with posttraumatic SCT and syringomyelia ASIA Score (mean ± SD) Testing Category* No. of Pts Pretest Posttest Difference p Value preop to 1st postop test motor strength light touch pinprick preop to 1 yr motor strength light touch pinprick preop to last FU motor strength light touch pinprick J Neurosurg: Spine / Volume 11 / October 2009

7 Posttraumatic spinal cord tethering and syringomyelia TABLE 5: The ASIA motor and sensory index scores in all 250 questionnaire respondents with posttraumatic SCT and syringomyelia No. of ASIA Score Testing Category* Pts Mean Median SD Min Max preop days from op motor strength light touch pinprick st FU days from op motor strength light touch pinprick yr FU days from op motor strength light touch pinprick last FU days from op motor strength light touch pinprick comparison of pre- and postoperative ASIA sensory and motor scores obtained at the longer mean follow-up intervals of 1 and 3.1 years failed to show significant differences. There was no association between the ASIA motor and sensory index outcomes and level of injury, age at surgery, years postinjury, sex, or need for cyst shunting. Fig. 3. Chart showing the pre- and postoperative testing results in a quadriplegic patient who became progressively dependent on a ventilator for 24 hours/day, subsequent to a progressive posttraumatic myelopathy from SCT. The testing scores in the outside columns represent motor testing at 18 months postoperatively. After surgery, this patient became ventilator independent for 10 hours/day at 6 months, for 24 hours/day at 18 months, and has maintained independence to the present (4 years). Note, however, that ASIA scoring does not reflect this functional improvement (pre- and postoperative motor scores of 0) or the improvement of tested accessory muscles of respiration. Note that this patient s pre- and postoperative MR imaging studies are shown in Fig. 4. J Neurosurg: Spine / Volume 11 / October 2009 Subjective Outcomes: Patient Questionnaire Responses The 250 patients who completed the questionnaire (Table 11) represented a mean follow-up duration of 6.15 years, with a range of years elapsed since study entry. Table 12 summarizes the questionnaire results for patients reporting motor and/or sensory loss as the reason for seeking neurosurgical treatment. Progressive loss of sensory and/or motor function was reported as a reason for seeking neurosurgical treatment by 204 of the 250 questionnaire respondents. Of these patients who presented with symptomatic sensory and/or motor deterioration, 89% believed that surgery resulted in arrest of the progressive functional loss, although 81% reported overall benefit from the surgery (Tables 11 and 12). Interestingly, 15% believed that postoperative sensory and/or motor function improved to a level greater than that achieved at any point postinjury, 19% believed that sensory and/or motor function returned to their best level achieved postinjury, and 12% believed that sensory and/or motor function returned partially back to their 451

8 S. P. Falci, C. Indeck, and D. P. Lammertse TABLE 6: The pre- to postoperative changes in ASIA motor and sensory index scores in all 250 questionnaire respondents with posttraumatic SCT and syringomyelia ASIA Score (mean ± SD) Testing Category* No. of Pts Pretest Posttest Difference p Value preop to 1st postop test motor strength light touch pinprick preop to 1 yr motor strength light touch pinprick preop to last FU motor strength light touch pinprick best level achieved postinjury, whereas 11% believed that surgery arrested progressive loss of function, although it decreased their functional state. In some, issues other than neurological loss contributed to this decrease in functional state. Although not specifically addressed in the questionnaire, some reported musculoskeletal issues such as postsurgical pain leading to disuse atrophy during the postoperative period. This was most common in the population of patients with quadriplegia. Of the patients completing the outcome questionnaire, 152 reported progressive loss of motor function preoperatively; 93% of these patients presenting with symptomatic motor deterioration believed that surgery arrested their progressive loss of motor function, 17% believed that surgery resulted in return of function to a level greater than that achieved at any time postinjury, 22% reported return of function to their greatest level postinjury, and 14% reported being partially restored to their greatest level of function postinjury, whereas 13% believed that surgery arrested progressive loss of function, but resulted in a decrease of their functional state (Table 12). Of the patients completing the questionnaire, 128 reported progressive loss of sensation prior to surgery; 97% of the 128 reported that surgery arrested progressive loss of function to the 11-year follow-up, 18% reported that surgery resulted in return of function to a level greater than that achieved at any time postinjury, 19% reported return of function to their greatest level postinjury, and 14% said they had return of function partially back to the greatest level achieved at any time postinjury, whereas 13% reported that surgery arrested progressive loss of function, but resulted in a decrease of their functional state (Table 12). Spasticity, Hyperhidrosis, Pain Ninety-nine (40%) of the 250 patients responding to the questionnaire reported progressively worsening spasticity prior to surgery. Fifty-nine (60%) of these patients presenting with worsening spasticity reported improvement postoperatively (Table 13). Fifty-three (21%) of the 250 patients responding to the questionnaire reported experiencing new or worsening hyperhidrosis prior to TABLE 7: The ASIA motor and sensory index scores in 152 questionnaire respondents presenting with motor loss related to posttraumatic SCT and syringomyelia ASIA Score Testing Category* No. of Pts Mean Median SD Min Max preop days from op motor strength light touch pinprick st FU days from op motor strength light touch pinprick yr FU days from op motor strength light touch pinprick last FU days from op motor strength light touch pinprick J Neurosurg: Spine / Volume 11 / October 2009

9 Posttraumatic spinal cord tethering and syringomyelia TABLE 8: The pre- to postoperative changes in ASIA motor and sensory index scores in 152 questionnaire respondents presenting with motor loss related to posttraumatic SCT and syringomyelia ASIA Score (mean ± SD) Testing Category* No. of Pts Pretest Posttest Difference p Value preop to 1st postop test motor strength light touch pinprick preop to 1 yr motor strength light touch pinprick preop to last FU motor strength light touch pinprick surgery, and 41 (77%) of these patients reported improvement of hyperhidrosis postoperatively. A small subset (38 patients) of the 362 eligible in the study underwent surgery for progressively worsening neuropathic pain as their only issue, and 47% of the patients in this subset reported that surgery resulted in relief of the pain to a degree that substantially improved the quality of their lives. In general, clinical correlation suggested subjective reduction in pain by at least 50%, although 100% pain relief was achieved in some cases (Table 13). With regard to the least commonly presenting symptoms of autonomic dysreflexia, hyperhidrosis, blurred vision, altered mentation, extremity edema, and skin changes, questionnaire responses were too small in number to assess meaningful long-term outcome. Regardless, long-term improvement of each of these presenting symptoms was reported. Need for Repeat Surgery A large majority (86.5%) of the 362 eligible patients underwent a single surgery during the study; 11.6% underwent a second surgery because of recurrent SCT, cyst (syrinx) reexpansion, CSF leakage, and/or pseudomeningocele formation; 1.6% of patients underwent 3 surgeries; and a single patient (0.3%) had 5 surgeries for these indications. Need for Shunt Tube Placement Of the 362 eligible patients in this study, 20% required cyst (syrinx) shunting in addition to detethering and duraplasty, based on the operative criteria described. Risks of Neurological Deterioration From Surgery For a better assessment of the risk of significant neurological loss from surgery, separate analyses were made of patients losing or gaining 3 ASIA motor points, and those losing or gaining 8 light touch or pinprick ASIA J Neurosurg: Spine / Volume 11 / October 2009 sensory points (Table 14). Because the majority of patients (68%) were quadriplegic, it was believed that a 3-point motor change would reflect functional significance. An TABLE 9: The ASIA motor and sensory index scores in 128 questionnaire respondents presenting with sensory loss related to posttraumatic SCT and syringomyelia ASIA Score Testing Category* No. of Pts Mean Median SD Min Max preop days from op motor strength light touch pinprick st FU days from op motor strength light touch pinprick yr FU days from op motor strength light touch pinprick last FU days from op motor strength light touch pinprick

10 S. P. Falci, C. Indeck, and D. P. Lammertse TABLE 10: The pre- to postoperative changes in ASIA motor and sensory index scores in 128 questionnaire respondents presenting with sensory loss related to posttraumatic SCT and syringomyelia ASIA Score (mean ± SD) Testing Category* No. of Pts Pretest Posttest Difference p Value preop to 1st postop test motor strength light touch pinprick preop to 1 yr motor strength light touch pinprick preop to last FU motor strength light touch pinprick point threshold was selected for sensory evaluation because it reflects 2 full dermatomes and probably would represent more than subjective testing error. With regard to motor loss/gain with testing shortly after surgery (> 90% within 2 weeks), 13% had motor loss of 3 points, 7% had motor gain of 3 points, and there were 87% with loss or gain of < 3 points. With motor testing on average at 1 year postoperatively, 9% had motor loss of 3 points, 14% had motor gain of 3 points, and there were 91% with loss or gain of < 3 points. With regard to light touch testing in the perioperative period (> 90% within 2 weeks), 10% had a loss of 8 points, 16% had a gain of 8 points, and 90% had a gain or loss of < 8 points. On average at 1 year postoperatively, 7% of patients had a loss of 8 points, 17% had a gain of 8 points, and 93% had a gain or loss of < 8 points. With regard to pinprick testing in the perioperative period (> 90% within 2 weeks), 7% had a loss of 8 points, 15% had a gain of 8 points, and 93% had a loss or gain of < 8 points. On average, at 1 year postoperatively, 12% had a loss of 8 points, 12% had a gain of 8 points, and 88% had a loss or gain of < 8 points. Postoperative Complications Of the 420 surgeries performed in the 362 eligible patients, 2 (0.48%) were treated for the postoperative complication of bacterial meningitis, 4 (0.95%) for pulmonary embolus, 6 (1.4%) for deep venous thrombosis, 16 (3.8%) for CSF leakage and/or pseudomeningocele, 2 (0.48%) for wound dehiscence/decubitus, and 1 (0.24%) for myocardial infarction. There were 2 deaths (0.48%), one from postoperative myocardial infarction and the other from meningitis (Table 15). Discussion In this study, we attempted the clinical evaluation of the surgical therapy of spinal cord detethering, expansion duraplasty, and when indicated, cyst (syrinx) shunting, to arrest a progressive myelopathy from posttraumatic SCT and its sequela of progressive cyst formation. Objective and subjective evaluations were used. Surgery overwhelmingly was offered for patients who perceived and experienced a substantial and progressive loss of sensory and/or motor function. Unfortunately, we were not able to correlate the experienced progressive loss with serial ASIA sensory and motor scoring performed preoperatively, because most patients underwent their initial rehabilitation at other centers and did not have routine scoring up to the time of surgery. We were, however, able to assess a large number of patients with serial ASIA scoring postoperatively, comparing their scores to a preoperative baseline. With this in mind, our objective results suggest that the surgical therapy we have described arrests a progressive myelopathy with regard to sensory and/or motor function, as assessed by ASIA scoring, for at least an average of years postoperatively, as long as only a single surgery is required (86% of patients in this study). The subtle decrease seen only in motor scoring at the same postoperative time frames, when multiple surgeries were factored in, suggests the possibility of greater resilience of sensory pathways to the negative influence of spinal cord retethering and cyst reexpansion. More than 90% of patients underwent their initial postoperative ASIA sensory/motor evaluation within 14 days of surgery. It is noteworthy that statistically significant improvements in light touch and pinprick, but not in motor scores, are seen in this early time frame. Because we maintain our patients on flat bed rest for 5 days postoperatively, with gradual mobilization subsequently, we anticipate disuse atrophy and weakness in these early time frames. Perhaps this accounts in part for a failure to see motor score improvement in this same period. That the sensory score improvement is not maintained at later 454 J Neurosurg: Spine / Volume 11 / October 2009

11 Posttraumatic spinal cord tethering and syringomyelia Fig. 4. A: A T2-weighted MR image of a posttraumatic spinal cord circumferentially tethered to the surrounding dura mater. Note regions of severe myelomalacia (double arrow). B: The spinal cord subsequent to surgical detethering. Note the substantial decrease of myelomalacia and reconstituted posterior subarachnoid space (double arrow). time frames perhaps suggests that a surgically detethered state achieves its maximal effect early. Questionnaire data allowed the assessment of patientreported long-term outcomes that extended the duration of follow-up to a mean of 6.15 years (range years). Our subjective results from this patient self-assessment suggest that surgery arrests a progressive myelopathy in 89% of those losing sensory function, motor function, or both, and provides functional return in 46%. Interestingly, 15% reported return of function to a level greater than that achieved at any time postinjury, suggesting that the tethering process and its sequelae can limit the magnitude of initial neurological recovery after injury. When one focuses specifically on the subset of patients losing motor function, patient self-assessment suggests that surgery results in functional arrest in 93%, functional return in 53%, and functional return to a level greater than that achieved at any time postinjury in 17%. When one focuses specifically on the subset of patients losing sensory function preoperatively, patient self- Fig. 5. A: A T2-weighted MR image of a posttraumatic tethered spinal cord with regions of descending cystic cavitation (large arrow), and ascending, predominantly precystic myelomalacia (small arrow). B: Collapse of the descending cyst (large arrow) and substantial resolution of the ascending precystic myelomalacia (small arrow) subsequent to spinal cord detethering, expansion duraplasty, and descending cyst subarachnoid shunting. Note that a shunt tube was not required for the ascending precystic myelomalacia. J Neurosurg: Spine / Volume 11 / October

12 S. P. Falci, C. Indeck, and D. P. Lammertse Fig. 6. A and C: Sagittal T2-weighted MR images of a posttraumatic tethered cervical spinal cord with a very large descending cyst (large arrows) and small region of ascending myelomalacia (small arrows). B and D: Sagittal MR images demonstrating marked collapse of the descending cyst (large arrows) and substantial resolution of the ascending myelomalacia (small arrow) subsequent to spinal cord detethering, expansion duraplasty, and descending cyst peritoneal shunting. assessment suggests that surgery results in arrest of functional loss in 97%, functional return in 51%, and return of function to a level greater than that achieved at any time postinjury in 18%. Although it was not evaluated in this study, it is our experience that return of function to a level greater than baseline is rarely clinically significant. It is noteworthy that the functional improvement reported by a substantial number of patients is not reflected in the ASIA scoring. As noted earlier, ASIA scoring is limited to evaluation of specific muscle groups for motor function, and specific points on the skin for sensory function. Although very useful, the limitations of ASIA sensory and motor scoring have become evident. With ASIA scoring, finger extension is not assessed; finger flexion is assessed but in a single muscle group of a single finger; and neck, shoulder girdle, deltoid, pectoral, and latissimus musculature are not assessed. Notably, the majority of patients in this study (68%) were quadriplegic and very dependent on the aforementioned muscle groups for activities such as typing, writing, feeding, grooming, bowel programs, dressing, wheeling, and body transfers. This disconnect between patient-reported functional improvement and ASIA scoring outcomes could be explained by the absence of assessment of these very muscle groups. Other functional improvements reported by patients that may not be reflected in ASIA motor scoring include improved muscular endurance and improved respiratory, bowel, bladder, and sexual function (Fig. 3). Risks related to surgery must not be understated. Subanalyses of the ASIA sensory/motor scoring data suggest that surgery may result in functional motor loss in 9%, light touch loss in 7%, and pinprick loss in 12%. These analyses are consistent with patient perception of functional loss subsequent to surgery (11% sensory and/ or motor). Although we assume that the majority of functional losses tested and reported by patients are from neurological causes, we recognize that a subset of patients may exist in whom functional loss may be a result of musculoskeletal issues, including postlaminectomy pain and disuse atrophy because of the pain, or delayed kyphosis caused by multilevel laminectomies. We have not quantified this subpopulation, however, to assess it clinically for such issues. Subjective self-assessment of postoperative outcome suggests that surgery improves spasticity in 60%, hyperhidrosis in 77%, and neuropathic pain in 47% when assessed at long-term follow-up. Other points are worthy of discussion. In our experi- 456 J Neurosurg: Spine / Volume 11 / October 2009

13 Posttraumatic spinal cord tethering and syringomyelia TABLE 11: Patient-focused outcome questionnaire results in 250 respondents with posttraumatic SCT and syringomyelia* % of Questions & Answers Pts 1. Why did you initially seek neurosurgical treatment? A. Loss of strength 61.8 B. Loss of sensation 51.2 C. Pain 66.0 D. Sweating 21.2 E. Change in spasticity W ere you aware of any motor or sensory changes prior to therapy testing? A. Yes 75.6 B. No 22.4 C. Don t know W ere specific activities affected (such as ability to transfer, hold a cup, or feed yourself)? A. Yes 65.9 B. No Do you feel the surgery resulted in A. 1. Return of lost function to baseline? 2. Return of lost function partially to baseline? 3. Return of function to greater than baseline? B. Prevention of further deterioration? C. P revention of further deterioration, but further loss of function? D. No effect on deterioration? 5. Co mpared to before your surgery, how would you rate your current status regarding Better Same Worse A. Pain? B. Spasticity? C. Sweating? D. Other? 6. Overall, do you feel you benefited from the surgery? A. Yes 81.0 B. No 16.0 C. Don t know 3.0 * The time from first surgery to questionnaire administration fell within the following parameters: mean 6.15 years; median 6.97 years; SD 3.05 years; minimum 0.93 years; maximum years. See Table 12 for results. See Table 13 for results. This question pertained to the 204 patients who sought surgical treatment. ence, injury to the spinal cord from any cause may result in tethering of the cord to the surrounding dura mater, and delayed neurological deterioration that may begin months to > 40 years subsequently. However, our experience suggests that SCT occurs in all cases of SCI related to skeletal trauma. Tethering may occur in cases of SCI unrelated to skeletal trauma, such as from hemorrhage, J Neurosurg: Spine / Volume 11 / October 2009 TABLE 12: Surgical results regarding functional status reported on patient-focused outcome questionnaire in 250 respondents with posttraumatic SCT and syringomyelia % of Status Preop & Results of Op Pts pts claiming progressive loss of sensory and/or motor function preop (total 204) arrest of progressive loss of function 89 im provement of sensory &/or motor function to better than 15 baseline improvement of sensory &/or motor function to baseline 19 im provement of sensory &/or motor function partially to 12 baseline function less than preop state 11 pts claiming progressive loss of motor function preop (total 152) arrest of progressive loss of motor function 93 improvement of motor function to better than baseline 17 improvement of motor function to baseline 22 improvement of motor function partially to baseline 14 function less than preoperative state 13 pts claiming progressive loss of sensation preop (total 128) arrest of progressive loss of sensory function 97 improvement of sensory function to better than baseline 18 improvement of sensory function to baseline 19 improvement of sensory function partially to baseline 14 function less than preoperative state 13 infection, previous surgery to the spinal cord, and prolonged compression from stenosis. In some cases of SCT, progressive myelomalacia and cyst expansion (syringomyelia) may additionally occur. In this series, significant cyst expansion, which was not adequately addressed by detethering and duraplasty, required shunting in 20% of the patients. In our experience, MR imaging is the best diagnostic modality for assessment of SCT, myelomalacia, and cyst formation, although CT myelography can be helpful if a metal artifact makes MR imaging nondiagnostic. Cine MR imaging may be useful, but is often difficult to obtain in patients with any degree of spasticity, because of movement artifacts. It should be noted that cine MR imaging does not assess CSF flow, but rather CSF motion, and normal results on a cine MR imaging study cannot definitively rule out localized flow abnormalities that indicate SCT. Patient history and physical examination findings are critically important in determining the relevance of the imaging findings of posttraumatic tethered, myelomalacic, and/or cystic spinal cord when making the diagnosis of a clinical progressive myelopathy. The progression of symptoms previously described must be correlated to objective clinical findings, and other potential contributory factors such as anatomical compression of the spinal cord, spinal instability, and primary neurodegenerative diseases such as multiple sclerosis should be ruled 457