18 Intramedullary Spinal Cord Tumors

Biopsy and Adjuvant Therapy

The optimal treatment of intramedullary spinal cord tumors (IMSCTs) has long been controversial; however, the evolving opinion of contemporary surgeons has been in favor of an aggressive surgical resection with or without postoperative adjuvant therapy.1–3

Technological advances in neuroimaging, microsurgical technique, neurophysiologic monitoring, and operative equipment have permitted neurosurgeons to undertake a more aggressive surgical approach, often achieving gross total resection (GTR) or subtotal resection (STR) based on intraoperative inspection of the tumor surgical site and postoperative magnetic resonance imaging (MRI).^4–27 Despite these advances, the risk of morbidity associated with aggressive surgical resection remains substantial, and the true benefit of aggressive surgical therapy has not been validated in terms of survival curves and functional outcomes scales when compared with less aggressive measures, such as biopsy for tissue diagnosis, along with adjuvant radiotherapy and/or chemotherapy.^8,28

Furthermore, the current body of literature supporting aggressive resection is based on retrospective series, case reports, and expert opinions. Class I data comparing aggressive surgical resection with tissue biopsy and adjuvant therapy are unavailable.

In this chapter we have been asked to support the use of a minimally aggressive surgical approach in the management of IMSCT. Data will be presented to support the point of view that biopsy followed as necessary by adjuvant therapy is a safe and effective way to manage these lesions and may be preferable to aggressive surgical excision.

Epidemiology

Intramedullary spinal cord tumors are rarely encountered neoplasms, particularly in the pediatric population, where the incidence is 4 to 10 per 10 million.^29,30 Among neoplasms of the central nervous system (CNS), IMSCTs account for 6 to 8% of all tumors.^6,29,31 In the pediatric population, astrocytomas are more prevalent, being 3 times more common than ependymomas, whereas the opposite holds true in the adult population.^9,32,33 Among pediatric astrocytomas, 25% are of malignant histology, being either anaplastic astrocytomas or glioblastoma multiforme.^1,34

Clinical Presentation

At the time of presentation, the most common findings are motor regression, pain, and gait abnormality. A smaller percentage may present with scoliosis or orthopedic abnormalities. In general, patients with more malignant lesions have a shorter prodrome of complaints before the correct diagnosis of spinal cord tumor is established.

Outcomes Analysis

To date, the body of literature supporting the aggressive surgical management of IMSCT is based on class IV evidence, comprised of uncontrolled studies, case series, case reports, and expert opinions.^{1,2,6–8,10,11,13,14,35–39}

The largest pediatric series to date on IMSCT is that by Constantini et al, which comprises the surgical series of Dr. Fred Epstein.¹ In this series, the authors retrospectively reviewed 164 children and young adults who underwent surgery for IMSCT. GTR was achieved in 76.8% and STR in 20.1%, with 79% of the lesions being histologically low grade. Over time there was an increasing bias toward use of neurophysiologic monitoring and laminoplasty. Pre- and postoperative functional evaluations were assessed with the Modified McCormick Scale (Table 18.1).¹ Twenty-three percent of patients worsened at least one functional grade after surgery, 60% were unchanged, and 17% improved. Patients with poor preoperative functioning had a greater risk of surgical morbidity, and in general, high-grade lesions were more disabled at presentation. Histologic analysis revealed low-grade tumors in 124, intermediate-grade in 19, and high-grade in 12 patients. Twenty-six percent of the group went on to have adjuvant therapy, including radiation and chemotherapy. Progression-free survival was estimated using the Kaplan-Meier technique. This revealed that GTR and STR were equally efficacious for low-grade tumors, and there was no survival benefit to GTR. Patients who underwent biopsy alone did fare worse, but these appear to have been almost exclusively high-grade lesions that behaved poorly regardless of attempted surgical approach. In fact, tumor histology was the only reliable predictor of patient survival. In summary, this study showed us that radical surgery can be performed in the majority of patients, but it did not prove that survival was superior to biopsy alone for low-grade lesions; furthermore, the patients were more likely to worsen than to improve their functional status with aggressive surgery.1

Dr. Epstein had a unique practice with an unusually high volume of patients with IMSCTs, and the data derived from his study are important to our understanding of these children. His work accurately describes the clinical course of pediatric patients undergoing aggressive resection of IMSCTs; however, a prospective, randomized control study or prospective matched group cohort study comparing aggressive surgical therapy with biopsy and adjuvant therapy has not been performed.

Radiotherapy

Many institutions reserve external beam radiotherapy for patients in whom GTR was not possible, in patients who have disease recurrence, or for high-grade lesions. Many authors have advocated aggressive surgery alone over subtotal resection or biopsy followed by radiation so as to avoid late complications of radiation. The current body of literature, however, does not necessarily support this rationale. In addition, in looking at the late effects of radiation, one must then equally consider the late effects of surgery on the pediatric spine to have an accurate comparison.

The largest, most comprehensive review of IMSCT in children treated at a single institution with surgery followed by external beam radiation was performed by O’Sullivan et al reviewing the data from the Hospital for Sick Children in Toronto.³⁴ Included in this study were 11 ependymomas and 15 astrocytomas, of which 12 were low grade, 3 high grade, and 5 other tumor types. Biopsy alone was performed in 35%, STR in 45%, and GTR in 19%. All patients received local high-dose radiation. The relapse rate based on the extent of resection was 37%, 14%, and 33%, for biopsy, STR, and GTR, respectively; thus, the extent of resection seemed to have no effect on the likelihood of recurrence. Local control of the tumor was achieved in 26 cases (84%), despite either grossly incomplete resection or biopsy alone in 25 of these cases (81%). Two patients did suffer from second malignancies that were felt to be radiation induced. Based on their results, the authors concluded that radiation treatment without resection may achieve long-term control in children with astrocytoma or ependymoma of the spinal cord, and in fact the results of progression-free survival were similar to the large surgical series of Constantini et al.¹It was more difficult to determine any differences in functional outcome between the two studies.^1,34 Of note, 68% of the patients did go on to suffer from progressive spinal deformity, therefore indicating that biopsy and adjuvant therapy may not be protective against deformity.³⁴ Other authors have shown that in low- and intermediate-grade astrocytomas, postoperative radiation therapy has improved overall survival or progression-free survival.^40–46

Chemotherapy

The role of chemotherapy in the management of IMSCTs has not been clearly defined. Its use has been reserved primarily as an adjuvant modality in cases of unresectable or recurrent lesions. There has been documentation in the literature of successful and inspiring treatment of intramedullary astrocytomas with chemotherapeutic agents; however, these have been isolated to case reports.^47–53

Postsurgical Spinal Deformity

The incidence of progressive spinal deformity following either laminectomy or laminoplasty for any indication has been previously reviewed.⁵⁴ Risk factors associated with increased incidence of deformity include the extent of facet resection, the number of lamina removed, the spinal segment involved, the presence of preoperative deformity, the growth potential of the spine (more common in pediatric patients), and IMSCTs.^55–67

The developing spine is especially susceptible to post-laminectomy deformity for several reasons. There is increased ligamentous laxity when compared with that of more mature skeletal systems. Furthermore, in the pediatric cervical spine, the orientation of the facet complex is more in the horizontal plane as compared with the vertical plane arrangement in the adult spine and therefore protects less against anterior subluxation.65,68 Finally, surgical disruption of the growth dynamics of the pediatric spine tends to propagate further deformities.^{55,56,64–66}

IMSCTs, in and of themselves, are known to be associated with sagittal plane deformity.^60,64,65 It has been postulated that involvement of the ventral horns may cause neuromuscular insufficiency, thus weakening the muscular support of the spine and leading to progressive spinal column deformity.^69,70 At the time of initial presentation, 15 to 40% of patients presenting with IMSCTs have an existing deformity, which is a known risk factor for progressive deformity following surgical resection of IMSCT.^61,64 Furthermore, when compared with those undergoing cervical decompression for degenerative conditions of the spine, patients with IMSCTs are twice as likely to develop deformity, with postoperative deformity being reported in 16 to 100% of patients in various studies.^56,57,71,72

Yao et al described the risk factors for progressive spinal deformity in children undergoing surgical resection of IMSCT: age younger than 13 years, the presence of preoperative deformity, and involvement of the thoracic or thoracolumbar segments increased the odds ratio of a subsequent spinal deformity requiring surgery for stabilization by 4.4, 3.2, and 2.6, respectively.⁶⁵ Although the extent of resection as an independent risk factor did not have a statistically significant predictive value on progression of deformity, children undergoing GTR (resection 95%) or STR (80–95% resection) had an odds ratio of 2.31 and 0.565, respectively, of developing postoperative deformity that required further surgery for stabilization, whereas those under going biopsy alone had an odds ratio of 0.001.⁶⁵

Efforts to reduce postoperative deformity by performing laminoplasty have shown marginal clinical reduction in the progression of sagittal imbalance.⁵⁴ Some authors advocate preemptive fusion at the time of the initial resection in patients who present with a preoperative kyphotic deformity.⁶⁸

Conclusion

Aggressive resection of IMSCTs carries a substantial risk of surgical morbidity, including neurologic decline and subsequent spinal column deformity requiring further surgical intervention for stabilization. In cases of ependymomas or low-grade astrocytomas, aggressive surgery may confer long-term survival benefits that outweigh the surgical risks, although this has not been established in any prospective study, and current data would indicate that the survival is similar to children undergoing biopsy and adjuvant radio-therapy. For patients with intermediate or high-grade lesions, surgical intervention does not appear to improve the overall survival, while subjecting these patients to high rates of surgical morbidity.

We respect the authors who have meticulously compiled and analyzed their experiences with IMSCTs. In light of the rarity of IMSCTs, however, it will require a multicenter, randomized, prospective study to formally address this controversial topic, especially for patients with low-grade lesions. Future studies should have specific emphasis on comparing survival, functional outcomes, and progression of spinal deformity. Currently, the literature does not define the optimal treatment algorithm for managing these complex patients, and the data supporting a conservative approach of initial biopsy followed by adjuvant therapy compare favorably to aggressive surgical approaches.

The Case for Surgery

Surgery for spinal cord tumors has a surprisingly long history, given the presumed difficulty of safely operating on the spinal cord. The first successful resection of an intramedullary spinal cord tumor was accomplished by Anton von Eiselsberg in Vienna in the fall of 1907.73 A two-stage surgical resection strategy (based on two cases) was developed shortly after that by Charles Elsberg in New York and published in 1911.⁷⁴

These bold pioneers had nothing of what is commonplace today: no useful imaging, no surgical technology, no micro-surgery, no monitoring, no neuroanesthesia. In subsequent decades, humanity was preoccupied with wars and economic disaster. There were no resources to further develop such seemingly “outlandish” surgeries. The neurologic risk of resecting intramedullary neoplasms was considered unacceptably high, and a conservative treatment concept was followed with a small role for surgery: an often timid biopsy, at best, combined with dural decompression to relieve the compartment-like compression of the swollen cord within the dural sac. Radiation therapy regardless of the histologic diagnosis was the factual treatment.75

Only in better times, and just prior to the advent of microsurgery, were new attempts made to remove tumors from the spinal cord.^76,77 Microneurosurgery reopened the door to a serious effort to establish surgical resection of intramedullary tumors as a management strategy superior to the biopsy–radiation concept. In the 1980s, after the microscope had established itself as an indispensable tool in neurosurgery, MRI dramatically improved the preoperative anatomical assessment of the spinal cord. Preoperative planning and postoperative follow-up became a reality. Thus, the conservative treatment strategy for intramedullary tumors gradually developed, and still continues to evolve, to a more active treatment concept with a larger role for aggressive surgical resection. This is aided by increased knowledge about the natural history of the disease and the response to surgery. Because most intramedullary tumors are low-grade neoplasms, complete and even near-complete resection appears to result in long-term progression-free survival with acceptable neurologic morbidity.^78–85 After the imaging revolution of the 1980s, the 1990s brought the development and application of intraoperative neurophysiologic monitoring. Understanding the functional integrity of the cord, particularly of the motor system, during surgery and in the process of neurologic recovery thereafter improved the surgeons’ ability to resect tumors with low neurologic morbidity.^86–89

Nevertheless, differences of opinion still exist about the proper referral of the relatively few children suffering from intramedullary tumors and their optimal individualized treatment. Despite the advances in surgical treatment, the old concept of “biopsy and radiation” is not extinct.^90,91 Lack of controlled evidence for a benefit of surgery has been suggested as an argument for a conservative treatment.

New drugs used for chemotherapeutic management of CNS tumors have not been widely used to treat intramedullary tumors. Only higher-grade glial tumors have been treated with a combination approach that includes chemotherapy.⁹² For low-grade tumors, little, if any, experience exists as to the impact of medical treatment of these neoplasms.⁹³ Given recent advances in the chemotherapeutic treatment of malignant gliomas of the brain,⁹⁴ it is reasonable to use the same regimen for glioblastomas of the spinal cord, and there may be further chemotherapy for spinal cord neoplasms in the future.

Epidemiology and Pathology

Intrinsic tumors of the spinal cord are rare. They comprise ~20 to 35% of all intradural neoplasms, with a higher percentage (55% of all intradural neoplasms) in children.⁹⁵ The most common intramedullary tumors in children are astrocytomas. Intramedullary ependymomas are the most frequent type in the adult age group but are rare exceptions in children.⁸⁵ Other tumors, such as hemangioblastomas⁹⁶ and cavernomas,⁹⁷occur.

The great majority of spinal cord tumors in children are benign, being either pilocytic or, less frequently, fibrillary astrocytomas. Like astrocytomas in the brain, the latter are poorly demarcated from normal tissue. Pilocytic tumors frequently have large areas of good demarcation, as well as cystic components, and smaller areas of diffuse growth next to the normal cord tissue.

Gangliogliomas are also low grade and occur primarily in children and young adults. Most frequently, intramedullary gangliogliomas grow slowly and thus have an indolent course and may present only when exceedingly large.⁹⁸

Ependymomas are the most common intramedullary neoplasm in adults,^99,100 whereas in children they account for only 12% of all intramedullary tumors.¹⁰¹ Typically, they have a central location in the spinal cord. Most often they are well circumscribed and clearly delineated from the surrounding spinal cord tissue. Practically all ependymomas are histologically benign.

Myxopapillary ependymomas are a subgroup of ependymomas with characteristic microcystic histologic features.¹⁰² They are usually located in the conus–cauda region.¹⁰³

Related posts:

Syringomyelia Chiari Malformations Slit Ventricle Syndrome Moyamoya Disease Craniopharyngioma Ependymomas

Stay updated, free articles. Join our Telegram channel

Join