Incidence of spinal deformity after resection of intramedullary spinal ...

J Neurosurg Pediatrics 1:57–62, 2008

Incidence of spinal deformity after resection of intramedullary spinal cord tumors in children who underwent laminectomy compared with laminoplasty MATTHEW J. MCGIRT, M.D.,1 KAISORN L. CHAICHANA, B.S.,1 APRIL ATIBA, B.S.,1 ALI BYDON, M.D.,1 TIMOTHY F. WITHAM, M.D.,1 KEVIN C. YAO, M.D.,2 AND GEORGE I. JALLO, M.D.1 Department of Neurosurgery, Johns Hopkins School of Medicine, Baltimore, Maryland; and Department of Neurosurgery, Tufts University, Boston, Massachusetts

1 2

Object. Gross-total resection of pediatric intramedullary spinal cord tumor (IMSCT) can be achieved in the majority of cases while preserving long-term neurological function. Nevertheless, postoperative progressive spinal deformity often complicates functional outcome years after surgery. The authors set out to determine whether laminoplasty in comparison with laminectomy has reduced the incidence of subsequent spinal deformity requiring fusion after IMSCT resection at their institution. Methods. The first 144 consecutive patients undergoing resection of IMSCTs at a single institution underwent laminectomy with preservation of facet joints. The next 20 consecutive patients presenting for resection of IMSCTs underwent osteoplastic laminotomy regardless of patient or tumor characteristics. All patients were followed up with telephone interviews corroborated by medical records for the following outcomes: 1) neurological and functional status (modified McCormick Scale [MMS] score and Karnofsky Performance Scale [KPS] score); and 2) development of progressive spinal deformity requiring fusion. The incidence of progressive spinal deformity and the long-term neurological function were compared between the laminectomy and osteoplastic laminotomy cohorts. The means are expressed 6 the standard deviation. Results. Overall, the patients’ mean age was 8.6 6 5 years, and they presented with median MMS scores of 2 (interquartile range [IQR] 2–4). A . 95% resection was achieved in 125 cases (76%). There were no differences (p . 0.10) between patients treated with osteoplastic laminotomy and those treated with laminectomy in terms of the following characteristics: age; sex; duration of symptoms; location of tumor; incidence of preoperative scoliosis (Cobb angle . 10°: 7 [35%] with laminoplasty compared with 49 [34%] with laminectomy); involvement of the cervicothoracic junction (7 [35%] compared with 57 [40%]); thoracolumbar junction (4 [20%] compared with 36 [25%]); tumor size; extent of resection; radiation therapy; histopathological findings; or mean operative spinal levels (7.5 6 2 compared with 7.5 6 3). Nevertheless, patients who underwent osteoplastic laminotomy had better median preoperative MMS scores than those treated with laminectomy (2 [IQR 2–2] compared with 2 [IQR 2–4]; p = 0.04). A median of 3.5 years (IQR 1–7 years) after surgery, only 1 patient (5%) in the osteoplastic laminotomy cohort required fusion for progressive spinal deformity, compared with 43 (30%) in the laminectomy cohort (p = 0.027). Adjusting for the intercohort difference in preoperative MMS scores, osteoplastic laminotomy was associated with a 7-fold reduction in the odds of subsequent fusion for progressive spinal deformity (odds ratio 0.13, 95% confidence interval 0.02–1.00; p = 0.05). The median MMS and KPS scores were similar between patients who underwent osteoplastic laminotomy and those in whom laminectomy was performed (MMS Score 2 [IQR 2–3] for laminotomy compared with 2 [IQR 2–4] for laminectomy, p = 0.54; KPS Score 90 [IQR 70–100] for laminotomy compared with 90 [IQR 80–90] for laminectomy, p = 0.545) at a median of 3.5 years after surgery. Conclusions. In the authors’ experience, osteoplastic laminotomy for the resection of IMSCT in children was associated with a decreased incidence of progressive spinal deformity requiring fusion but did not affect long-term functional outcome. Laminoplasty used for pediatric IMSCT resection may decrease the incidence of progressive spinal deformity requiring subsequent spinal stabilization in some patients. (DOI: 10.3171/PED-08/01/057)

KEY WORDS • intramedullary spinal cord tumor • laminectomy • laminoplasty • osteoplastic laminotomy • pediatric neurosurgery • spinal deformity

Abbreviations used in this paper: CSF = cerebrospinal fluid; IMSCT = intramedullary spinal cord tumor; IQR = interquartile range; KPS = Karnofsky Performance Scale; MMS = modified McCormick Scale; MR = magnetic resonance.

J. Neurosurg.: Pediatrics / Volume 1 / January 2008

NTRAMEDULLARY spinal cord tumors comprise approximately 35% of all pediatric intraspinal tumors.2,30 Historically, management of these tumors has involved biopsy sampling, dural decompression, and radiation therapy. However, with recent advances in microsurgical tech-

I

57

M. J. McGirt et al. nology, management of these tumors has increasingly shifted toward aggressive treatment with radical resection.3,4,14–16 This approach has been shown to be associated with increased long-term survival and improved quality of life.3,4, 13,15,28 Nevertheless, postoperative progressive spinal deformity often develops and complicates functional outcomes years after surgery.5,17,18,22,32 Spinal deformity is a well-documented complication following IMSCT resection, and has been reported in 24–100% of cases.5,17,18,22,32 This high incidence is probably multifactorial.9 Some studies have reported that the length and extent of laminectomy, which leads to a loss of posterior supporting elements, is directly associated with an increased risk of postoperative deformity.17,22,32 As a result, some investigators have theorized that osteoplastic laminotomy, which permits the replacement of bone, is associated with less spinal deformity.5,12,23 However, in a 2003 metaanalysis, investigators found that the rate of deformity following osteoplastic laminotomy was comparable to the rate of postlaminectomy deformity.24 Clinical studies favoring osteoplastic laminotomy over laminectomy for IMSCT resection still are lacking. In this retrospective review we set out to determine whether osteoplastic laminotomy compared with laminectomy has reduced the incidence of subsequent spinal deformity requiring fusion after IMSCT resection at our institution. Clinical Material and Methods As previously described,31 the records of 164 consecutive patients with IMSCTs surgically treated between 1980 and 1994 at a single academic institution were retrospectively reviewed for the occurrence of progressive spinal deformity requiring fusion. The first 144 patients underwent laminectomy with preservation of facet joints for access to the IMSCT, regardless of patient characteristics. The last 20 consecutive patients presenting for resection of their IMSCT underwent laminoplasty regardless of patient or tumor characteristics. Patient Population

One hundred sixty-four patients underwent resection of IMSCT during the study period. One hundred (61%) were male, and the mean age was 8.6 6 5.7 years. (The means are expressed 6 the standard deviation throughout.) Overall, the median preoperative MMS score19 was 2 (IQR 2–4). Symptoms were present for , 1 month prior to surgery in 24 patients (15%) and . 12 months in 47 patients (29%). One hundred forty-nine patients (91%) presented with motor symptoms, 120 (74%) with sensory symptoms, 53 (32%) with urinary incontinence, and 56 (34%) with stable preoperative scoliosis (Cobb angle . 10°). The tumors involved a mean of 6 6 3 levels. One hundred eighteen patients (72%) had a tumor-associated syrinx on MR images. Twenty-four tumors (15%) were confined to the cervical spine, 42 (25%) to the thoracic spine, 64 (39%) involved the cervicothoracic spine, and 40 (24%) the thoracolumbar spine. Gross-total resection was performed in 125 patients (76%) and subtotal resection in 33 (20%). Six patients (4%) underwent open biopsy procedures only. The mean number of laminectomies or laminoplasty levels was 7.5 6 3. Pathologically, 66 (40%) of the lesions were Grade I or II, 14 (9%) were Grade III, and 4 (2%) were Grade IV astrocytoma. In addition, there were 44 gangliogliomas (27%), 19 ependy58

momas (12%), 10 ganglioneurocytomas (6%), 3 glioneurofibromas (2%), 3 myxopapillary ependymomas (2%), and 1 primitive neuroectodermal tumor (1%). For the creation of the institutional database, all clinical and radiological variables at presentation were recorded. Patients underwent pre- and postoperative MR imaging in all cases. Each patient’s neurological examination was documented at hospital discharge and at a 3-month follow-up visit. For long-term follow-up evaluation, patients and their families were contacted by phone by a trained nurse practitioner and completed a questionnaire assessing symptoms, functional status, and subsequent medical occurrences. The medical records of patients were then obtained to confirm the answers in patient questionnaires and phone interviews. Assessment of outcomes did not occur at a unified point in time at the end of the study period. Patients were contacted as described at various times over these 14 years. Hence, there was a rolling follow-up assessment. On average, we were able to maintain contact for the purpose of this database for 3–4 years for almost all patients. Therefore, all patients not requiring surgery for deformity were eventually lost to follow-up at varying times (median 3.5 years). The documented detailed functional status was retrospectively graded according to an MMS19 (Table 1), and the KPS score.27 During the follow-up period, patients demonstrating continued progression of structural kyphotic or scoliotic curves on radiographs despite external bracing underwent attempted curve correction and instrumented fusion. Patients with stable spinal deformity underwent attempted curve correction and fusion only if their condition was accompanied by radicular symptoms, significant back pain, or functional limitation thought to arise from their spinal deformity. In children requiring subsequent fusion for progressive spinal deformity, the curve type (kyphosis, scoliosis, or kyphoscoliosis) and duration between prior IMSCT resection and fusion were documented. This institutional database was retrospectively reviewed for the purposes of this study. Surgical Technique

For laminoplasty, only medial facet joint exposure is performed by subperiosteal paraspinal muscle dissection. An effort was made to preserve the facet joint capsules in all cases. The interspinous ligaments at the rostral-most and caudal-most levels of the laminoplasty segment were removed with a Leksell rongeur. The spinous processes of the planned laminoplasty segment were left intact to preserve TABLE 1 Modified McCormick Scale for functional evaluation of patients with IMSCTs Grade

I II III IV V

Explanation

neurologically intact, ambulates normally, may have minimal dysesthesia mild motor or sensory deficit, patient maintains functional independence moderate deficit, limitation of function, independent w/ external aid severe motor or sensory deficit, limit of function w/ a dependent patient paraplegia or quadriplegia, even if there is flickering movement


Spinal deformity after IMSCT resection

FIG. 1. Illustrations depicting laminoplasty with ultrasonic aspiration of an IMSCT in an inside-out method until the interface with the white matter is reached (A–C). Tumor-associated syrinx is not exposed or surgically manipulated due to frequent spontaneous improvement after tumor resection alone. After tumor resection and primary dural closure, the laminae are reapproximated and attached with titanium microplates (D). CUSA = Cavitron ultrasonic surgical aspirator.

the interspinous ligaments and ligamentum flavum. Small interlaminar fenestrations and partial laminotomies were then made with a 2-mm Kerrison punch at the caudal edge of the caudal-most lamina, allowing a purchase groove for the osteotome footplate and identification of the dura mater before drilling. Bilateral laminotomies were then made in a caudal-to-rostral direction spanning the entire laminoplasty segment in a continuous cut. The laminoplasty width was made equal to the spinal canal. The ligamentum flavum was then removed with a 1-mm Kerrison rongeur and the laminoplasty segment was lifted with curettes, exposing the dura mater. The tumor was approached and removed via a midline dural incision spanning the entire length of the lesion. Following tumor resection and primary dural closure, the laminae were reapproximated and attached with titanium microplates (Fig. 1). Last, the paraspinal muscles were reapproximated to the laminae and sutured to the deep interspinous ligaments. In children undergoing laminectomy the laminae were removed the entire length of the tumor. Only medial facet joint exposure was performed in all cases. An effort was J. Neurosurg.: Pediatrics / Volume 1 / January 2008

made to preserve the facet joint capsules in all cases. Facetectomy was not performed in any patient treated with laminectomy. Tumor-associated syrinx was not exposed or surgically manipulated in any case due to its frequent resolution after tumor resection alone. Ultrasonic aspiration was always used to excavate the tumor from the inside outward until its interface with the white matter was reached. Sensory evoked and motor evoked potentials were used. Resection was deemed complete once interface with the white matter was reached or if a sustained decrease in motor evoked potentials . 50 % of baseline occurred. Gross-total removal of the tumor was attempted in most patients. Cases with no residual enhancement noted on postoperative MR imaging were classified as gross-total resection. Statistical Analysis

The incidence of spinal fusion for progressive spinal deformity, MMS scores, and KPS scores at last follow-up were compared between the laminectomy and laminoplasty cohorts. For intergroup comparison, the Student t-test was 59

M. J. McGirt et al. TABLE 2 Comparison of laminoplasty and laminectomy for resection of IMSCTs in 164 children* Variable

no. of patients clinical male mean age in yrs, 6 SD median preop MMS score symptom duration .1 yr radiological finding preop scoliosis† cervicothoracic junction thoracolumbar junction tumor-associated syrinx treatment mean surgical levels, 6 SD gross-total resection postop radiotherapy histological finding astrocytoma ganglioglioma other‡

Laminoplasty (%) Laminectomy (%) p Value

20 14 (70) 8.2 6 5 2 [2–2] 6 (30)

144 86 (60) 7.2 6 5 2 [2–4] 41 (28)

Outcome

0.622 0.222 0.048 0.947

7 (35) 7 (35) 4 (20) 13 (65)

49 (34) 57 (40) 36 (25) 105 (73)

0.916 0.808 0.784 0.415

7.5 6 2 14 (70) 6 (30)

7.5 6 3 111 (77) 36 (25)

0.773 0.387 0.786

73 (51) 38 (26) 33 (23)

0.781 0.792 0.423

11 (55) 6 (30) 3 (15)

* Patients who underwent laminoplasty had slightly less neurological deficit based on preoperative MMS scores. Otherwise, treatment cohorts were similar. The numbers in brackets represent the IQR. Abbreviation: SD = standard deviation. † Cobb angle . 10°. ‡ Category includes ependymoma, ganglioneurocytoma, glioneurofibroma, myxopapillary ependymoma, and primitive neuroectodermal tumor.

used for parametric data and the Mann–Whitney U-test for nonparametric data. Percentages were compared using chisquare tests or the Fisher exact test, depending on sample size. The independent association of laminoplasty or laminectomy with the incidence of progressive spinal deformity was assessed using multivariate logistic regression analysis (Statview, SAS Institute, Inc.) adjusting for all covariates differing between cohorts (preoperative MMS score). Results Laminoplasty Compared With Laminectomy

One hundred forty-four patients underwent laminectomy for IMSCT resection and 20 underwent laminoplasty. Patients undergoing laminoplasty had a lower preoperative MMS score compared with laminectomy patients (Table 2). Otherwise, there were no differences between treatment groups. Overall, the median follow-up duration was 3.5 years (IQR 1–7 years) and it was similar between cohorts (Table 3). No patient in the laminoplasty cohort developed an incisional CSF leak or pseudomeningocele. Four patients (3%) in the laminectomy group experienced pseudomeningocele formation or an incisional CSF leak. Only 1 patient (5%) in the laminoplasty cohort required fusion for progressive spinal deformity by the last follow-up evaluation. This patient developed a progressive scoliotic deformity of the thoracic spine 6 years after undergoing a 6-level thoracic laminoplasty as a 1-year-old child. A greater number of patients (43 [30%]) in the laminectomy cohort required fusion by the last follow-up evaluation for progressive spinal deformity (15 kyphosis, 14 scoliosis, and 14 kyphoscoliosis; p = 0.027; Table 3 and Fig. 2). Adjusting for differences in pre60

TABLE 3 Comparison of neurological function, performance status, and incidence of spinal fusion for progressive spinal deformity in patients undergoing osteoplastic laminotomy versus laminectomy for resection of pediatric IMSCTs* Laminoplasty

Laminectomy

no. of patients 20 144 median MMS score at 3 mos† 3 [2–3] 3 [2–4] median duration from last FU† 3.2 [2–5] yrs 3.6 [1–7] yrs median MMS score at last FU† 2 [2–3] 2 [2–4] median KPS score at last FU† 90 [70–100] 90 [80–90] patients w/ progressive 1 (5%) 43 (30%) deformity by last FU

p Value

0.480 0.254 0.541 0.545 0.027

* FU = follow-up. † Numbers in brackets denote the IQR.

operative MMS score for a multivariate analysis, laminoplasty remained associated with a 7-fold reduction in odds of subsequent fusion for progressive spinal deformity (odds ratio 0.13, 95% confidence interval 0.02–1.00; p = 0.05). The MMS and KPS scores were similar between osteoplastic laminotomy and laminectomy patients 3 months after surgery and at last follow-up (Table 3). Discussion In this study, we compared the impact of laminectomy with that of laminoplasty on the development of spinal deformity requiring fusion following IMSCT resection in children. In this series of 164 patients, the first 144 consecutive patients underwent laminectomy with preservation of the facet joints, and the next 20, regardless of patient or tumor characteristics, underwent laminoplasty. At a median duration of 3.5 years following surgery, patients who underwent laminoplasty compared with those who underwent laminectomy demonstrated a reduction in the development of progressive spinal deformity requiring fusion, but had similar functional status. This association persisted despite adjusting for the only patient characteristic that differed between treatment groups (preoperative MMS score). Adults with normal preoperative spinal alignment rarely develop postoperative spinal deformity following IMSCT resection.8,9,28 Pediatric patients, however, are predisposed to

FIG. 2. Kaplan–Meier curves demonstrating fusion as a function of time after surgery. A lower percentage of patients in the laminoplasty group required fusion compared with the laminectomy group (p , 0.05, log-rank analysis). Only a single patient required fusion for spinal deformity after laminoplasty (fusion was performed 6 years after tumor resection).


Spinal deformity after IMSCT resection developing spinal deformity,3,4,9 with rates varying between 16 and 100% in several series.5,10,22,25,32 This predisposition is probably multifactorial. Pediatric patients have a greater proportion of cartilage in their immature skeletal system, and thus increased ligamentous laxity.9 Additionally, they have more horizontally oriented facet complexes, compared with the vertically oriented complexes seen in adults.9 These features, in addition to their growing spine, accentuate the altered spinal biomechanics caused by IMSCT resection,9,26 leading to increased incidences of spinal deformity. Laminectomy has long been considered the traditional approach for IMSCT resection because it is presumably well established, creates a relatively wide exposure of the spinal cord, and can be easily extended in the rostral or caudal directions.17 In 1911, Elsberg and Beer6 described a 2-stage procedure in which laminectomy and myelotomy were performed at the initial operation, followed by IMSCT removal. Since then, several other studies have used this same approach.3,4,7,8,13,15,28 However, as the survival increased for children with IMSCTs, so did the incidences of spinal instability. Papagelopoulos et al.22 reported on the development of spinal deformity in 33% of children undergoing laminectomy for an IMSCT. Furthermore, de Jonge et al.5 described the development of postlaminectomy deformity in 67 (88%) of 76 patients with IMSCTs. Laminoplasty has therefore been advocated to avoid such complications, because the posterior elements of the spinal cord are replaced after en bloc removal of the laminae.1,12,23 This replacement is presumed to leave the posterior element intact, theoretically stabilizing the spine and preventing instability. A theoretical benefit in reducing CSF leakage or pseudomeningocele formation has been suggested to occur with laminoplasty. Although this added laminar barrier may help contain CSF within the spinal canal, the small difference observed in our study (0% compared with 3%) does not offer conclusive evidence to this benefit. This procedure was first described in 1976 by Raimondi et al.23 Since then, osteoplastic laminotomy has been used for several conditions, including cervical myelopathy, tethered spinal cord, syringomyelia, spasticity, and spinal tumors, among others.1,11,12,20,29 Biomechanical and clinical studies have provided inconclusive data supporting the use of laminoplasty over laminectomy.9,21,24,32,33 Nowinski et al.,21 using cadaveric specimens, found that spinal instability was increased after laminectomy when at least 25% of the facet joints were removed, compared with laminoplasty. Yeh et al.32 retrospectively found that laminoplasty was associated with decreased spinal deformity in patients undergoing IMSCT resection in the thoracolumbar region. The extent of facetectomy in the Yeh study was unclear, and was limited to a comparison of 9 patients undergoing laminectomy and 6 patients undergoing laminoplasty. However, in a 2003 metaanalysis, investigators found high rates of deformity following osteoplastic laminotomy (~ 35%) that were comparable to the reported rates of deformity following laminectomy.24 Prior studies therefore have not yet established a conclusive efficacy of laminoplasty over laminectomy in preventing spinal deformity after IMSCT resection in children. In our study, all surgeries were performed by the senior authors (F Epstein and GI Jallo) with standard surgical procedures and facet preservation for all patients. Furthermore, our high volume of patients and long-term follow-up J. Neurosurg.: Pediatrics / Volume 1 / January 2008

strengthens our cohort comparison. Our complete switch to laminoplasty regardless of patient characteristics eliminates treatment bias, which was highlighted by the similarities between the 2 cohorts. Nevertheless, our study is subject to the inherent bias associated with all retrospective studies. Whereas the neurological examination and functional status were documented in detail for inclusion in the IMSCT database, retrospective classification of the MMS grade weakens its accuracy, but does so equally between treatment cohorts. With a median follow-up duration of 3.5 years, the true incidence of postoperative deformity may be underestimated. Furthermore, although the association of laminoplasty and decreased deformity was observed within the first 4 years after surgery, our follow-up period does not allow this presumed effect to be applied to longer durations after surgery. Longer follow-up studies are needed to determine if this observed beneficial effect of laminoplasty persists 10 and 15 years after tumor resection. Last, although the percentage of patients with preoperative deformity did not differ between treatment groups, the degree of preoperative deformity was not accounted for in this study design. Patients with more severe preoperative deformity were most often scheduled for planned curve correction and were not included in this study. Nevertheless, the potential for variance in degree of preoperative deformity as a bias in this study cannot be ruled out. Conclusions In our experience, laminoplasty for the resection of IMSCT in children was associated with a decreased incidence of progressive spinal deformity requiring fusion, but did not affect long-term functional outcome. Laminoplasty for pediatric IMSCT resection may decrease the incidence of progressive spinal deformity and prevent the need for subsequent spinal stabilization in pediatric patients within the first 4 years after surgery. Acknowledgments We recognize the late Fred Epstein, M.D., for his role as senior surgeon in the care of all patients included in this study. We also thank Ian Suk for his artistic contribution to this manuscript. References 1. Abbott R, Feldstein N, Wisoff JH, Epstein FJ: Osteoplastic laminotomy in children. Pediatr Neurosurg 18:153–156, 1992 2. Barker DJ, Weller RO, Garfield JS: Epidemiology of primary tumors of the brain and spinal cord: a regional survey in southern England. J Neurol Neurosurg Psychiatry 39:290–296, 1976 3. Constantini S, Houten J, Miller DC, Freed D, Ozek MM, Rorke LB, et al: Intramedullary spinal cord tumors in children under the age of 3 years. J Neurosurg 85:1036–1043, 1996 4. Constantini S, Miller DC, Allen JC, Rorke LB, Freed D, Epstein FJ: Radical excision of intramedullary spinal cord tumors: surgical morbidity and long-term follow-up evaluation in 164 children and young adults. J Neurosurg 93 (2 Suppl):183–193, 2000 5. de Jonge T, Slullitel H, Dubousset J, Miladi L, Wicart P, Illés T: Late-onset spinal deformities in children treated by laminectomy and radiation therapy for malignant tumors. Eur Spine J 14:765–771, 2005 6. Elsberg CA, Beer E: The operability of intramedullary tumors of the spinal cord. A report of two operations with remarks upon

61

M. J. McGirt et al.

7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

62

the strusion of intraspinal tumors. Am J Med Sci 142:636–647, 1911 Epstein F, Epstein N: Surgical management of holocord intramedullary spinal cord astrocytomas in children. J Neurosurg 54:829–832, 1981 Epstein FJ, Farmer JP, Freed D: Adult intramedullary astrocytomas of the spinal cord. J Neurosurg 77:355–359, 1992 Fassett DR, Clark R, Brockmeyer DL, Schmidt MH: Cervical spine deformity associated with resection of spinal cord tumors. Neurosurg Focus 20(2):E2, 2006 Fraser RD, Paterson DC, Simpson DA: Orthopaedic aspects of spinal tumors in children. J Bone Joint Surg Br 59:143–151, 1977 Heller JG, Edwards CC II, Murakami H, Rodts GE: Laminoplasty versus laminectomy and fusion for multilevel cervical myelopathy: an independent matched cohort analysis. Spine 26:1330–1336, 2001 Hukuda S, Ogata M, Mochizuki T, Shichikawa K: Laminectomy versus laminoplasty for cervical myelopathy: brief report. J Bone Joint Surg Br 70:325–326, 1988 Jallo GI, Danish S, Velasquez L, Epstein F: Intramedullary lowgrade astrocytomas: long-term outcome following radical surgery. J Neurooncol 53:61–66, 2001 Jallo GI, Freed D, Epstein F: Intramedullary spinal cord tumors in children. Childs Nerv Syst 19:641–649, 2003 Jallo GI, Freed D, Epstein FJ: Spinal cord gangliogliomas: a review of 56 patients. J Neurooncol 68:71–77, 2004 Jallo GI, Kothbauer KF, Epstein FJ: Intrinsic spinal cord tumor resection. Neurosurgery 49:1124–1128, 2001 Katsumi Y, Honma T, Nakamura T: Analysis of cervical instability resulting from laminectomies for removal of spinal cord tumor. Spine 14:1171–1176, 1989 Lunardi P, Licastro G, Missori P, Ferrante L, Fortuna A: Management of intramedullary tumors in children. Acta Neurochir (Wien) 120:59–65, 1993 McCormick PC, Torres R, Post KD, Stein BM: Intramedullary ependymoma of the spinal cord. J Neurosurg 72:523–532, 1990 Meyer NJ, Flatley TJ, Dunn DD: Superiorly based laminoplasty in children: average 6.8-year follow-up of 21 patients. J Spinal Disord Tech 16:156–162, 2003 Nowinski GP, Visarius H, Nolte LP, Herkowitz HN: A biomechanical comparison of cervical laminaplasty and cervical laminectomy with progressive facetectomy. Spine 18:1995–2004, 1993 Papagelopoulos PJ, Peterson HA, Ebersold MJ, Emmanuel PR,

23. 24. 25. 26.

27. 28.

29. 30. 31.

32. 33.

Choudhury SN, Quast LM: Spinal column deformity and instability after lumbar or thoracolumbar laminectomy for intraspinal tumors in children and young adults. Spine 22:442–451, 1997 Raimondi AJ, Gutierrez FA, Di Rocco C: Laminotomy and total reconstruction of the posterior spinal arch for spinal canal surgery in childhood. J Neurosurg 45:555–560, 1976 Ratliff JK, Cooper PR: Cervical laminoplasty: a critical review. J Neurosurg 98 (3 Suppl):230–238, 2003 Reimer R, Onofrio BM: Astrocytomas of the spinal cord in children and adolescents. J Neurosurg 63:669–675, 1985 Saito T, Yamamuro T, Shikata J, Oka M, Tsutsumi S: Analysis and prevention of spinal column deformity following cervical laminectomy. I. Pathogenetic analysis of postlaminectomy deformities. Spine 16:494–502, 1991 Schag CC, Heinrich RL, Ganz PA: Karnofsky performance status revisited: reliability, validity, and guidelines. J Clin Oncol 2:187–193, 1984 Shrivastava RK, Epstein FJ, Perin NI, Post KD, Jallo GI: Intramedullary spinal cord tumors in patients older than 50 years of age: management and outcome analysis. J Neurosurg Spine 2: 249–255, 2005 Spiegel DA, Loder RT, Alley KA, Rowley S, Gutknecht S, Smith-Wright DL, et al: Spinal deformity following selective dorsal rhizotomy. J Pediatr Orthop 24:30–36, 2004 Steinbok P, Cochrane DD, Poskitt K: Intramedullary spinal cord tumors in children. Neurosurg Clin N Am 3:931–945, 1992 Yao KC, McGirt MJ, Chaichana KL, Constantini S, Jallo GI: Risk factors for progressive spinal deformity following resection of intramedullary spinal cord tumors in children: an analysis of 161 consecutive cases. J Neurosurg 107 (6 Suppl Pediatrics):463–468, 2007 Yeh JS, Sgouros S, Walsh AR, Hockley AD: Spinal sagittal malalignment following surgery for primary intramedullary tumors in children. Pediatr Neurosurg 35:318–324, 2001 Zdeblick TA, Zou D, Warden KE, McCabe R, Kunz D, Vanderby R: Cervical stability after foraminotomy. A biomechanical in vitro analysis. J Bone Joint Surg Am 74:22–27, 1992

Manuscript submitted July 15, 2007. Accepted September 14, 2007. Address correspondence to: Matthew J. McGirt, M.D., 3553 Newland Road, Baltimore, Maryland 21218. email: mmcgirt1@jhmi.edu.
