Disclosure
Intrinsic tumors of the spinal cord are rare neoplasms that can leave patients neurologically and functionally devastated. They have been diagnosed and surgically addressed for more than 100 years, often with limited success. Recently, improved neuroimaging has allowed these lesions to be diagnosed with accuracy. However, several advances in surgical technique were required to allow definitive treatment of these tumors. History of the Procedure: Important events in the development of our understanding of spinal cord tumors
Problem: Spinal tumors can be divided into extradural, intradural extramedullary, and intradural intramedullary. Frequency:
Etiology: A tumor can arise from any component of the spinal cord. The etiology of these tumors generally is unknown, although some may be familial (see hemangioblastoma in Picture 1). More than 90% are benign and subject to potential resection (see Picture 5). Clinical:
Examination
Relevant Anatomy: The spinal cord originates at the foramen magnum and extends to the conus medullaris (see Picture 1), which terminates at the L1-L2 vertebral body junction in adults. The pia matter condenses caudally to this area, extending to the sacrum as the filum terminale. It is an ovoid structure, most narrow in the anterior-posterior direction. Thirty-one pairs of spinal nerves exit from the lateral aspect. In contrast to the brain, the white matter of the spinal cord surrounds the interior gray matter. The spinal cord is surrounded by a supporting connective tissue layer of pia matter. A vestigial extension of the ventricular system, the central canal runs the length of the spinal cord. Dilation of the central canal may be pathologic in some instances. Cell types are identical to those of the brain. The vascular supply is from multiple sources. The vertebral arteries form the anterior spinal artery, which irrigates the anterior cervical and upper thoracic cord. Thirty-one spinal radicular arteries supply the remainder of the cord. The most important radicular artery is the artery of Adamkiewicz. This large vessel occurs in variable locations but most often on the left side, from T9 to L2. In general, it should be preserved during surgical procedures. T1-T4 and L1 are vulnerable areas for vascular insult due to decreased interconnections of the vascular supply. Contraindications: Relative contraindications to surgical intervention include complete neurologic deficit over 24 hours and short life expectancy.
Lab Studies:
Imaging Studies:
Other Tests:
Diagnostic Procedures:
Ependymoma (56%) The most common intrinsic spinal cord tumor has a male predilection and a fourth-decade prevalence. They occur anywhere in the cord and are commonly in the conus medullaris, where an exophytic component may be present. They rarely change growth characteristics and metastasize. Lesions are characteristically hypovascular, well circumscribed, and noninfiltrative of the surrounding cord. Symptoms are due to compression of the surrounding cord rather than infiltration. Complete resection often results in prolonged survival. Astrocytoma (29%) These lesions are more common in children than in adults. Sometimes they are associated with microcysts or syrinxes. The pilocytic varieties are well differentiated and tend to be indolent, with a definable surgical plane. The remainder of low-grade astrocytomas have well-defined planes in approximately 30-50% of cases, making complete resection difficult at times. Residual tumor often has an indolent course, and controversy exists in the management of such tumors. Fortunately, anaplastic astrocytoma or glioblastoma are rare. These malignant tumors exhibit rapid growth, are locally invasive, and may seed the CSF. Distinguishing between tumor and normal cord is difficult. Surgical therapy does not improve the dismal course, with death usually occurring within 2 years. Oligodendroglioma (3%) Knowledge regarding this rare spinal cord neoplasm is limited. Whereas the intracranial type of this tumor has exhibited responses to chemotherapy, its use is unproven in this setting. Developmental tumors (3%) Dermoid, epidermoid, and teratoma are slow-growing neoplasms with a thoracolumbar predominance. Some dermoids of the conus medullaris have been attributed to lumbar punctures that carry in cutaneous tissue. These may have a dense capsule, precluding complete removal; although, this may be compatible with prolonged symptom-free survival. When complete removal is unobtainable, debris produced by the tumor may cause an early recurrence of symptoms. Hemangioblastoma (3%) This is a vascular tumor that is associated with von Hippel-Lindau disease in 30% of cases. It may have an associated syrinx and occur in multiple locations. These should not be removed in a piecemeal fashion because significant bleeding may ensue, increasing the risk of the procedure. Removal of the lesion is considered curative. Lipoma (2%) Not true neoplasms, they present in the first 3 decades of life when fat is being deposited. They may be associated with cutaneous abnormalities. Loss of total body fat may be necessary to reduce the mass of the tumor. Fibrous adhesions to the cord make total removal difficult. Removal is not the goal of surgery. The carbon dioxide laser is particularly useful during surgery for this lesion. Others (4%) Unusual lesions include subependymoma, ganglioglioma and intramedullary schwannoma, and neurofibroma. Management of low-grade lesions parallels other indolent lesions. Metastatic lesions to the spinal cord are unusual. Large series defining the management of these tumors are not available. |
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Medical therapy: Because the majority of these tumors are slow growing and locally contained, surgical extirpation, where possible, is the treatment of choice. In selected situations, watchful waiting can be considered. Steroids are used in the perioperative period or if a rapid decline in neurologic function occurs, but steroids are not considered tumoricidal. Radiation therapy The slow-growing nature of these neoplasms makes proving the benefit of this treatment difficult. Conclusions regarding the efficacy of radiation therapy as a primary therapy are not available for all tumor types. Series have shown poor control of local disease in ependymomas. Data are available that suggest surgically excised ependymomas need not undergo subsequent radiation therapy. Evidence of this modality preventing recurrence or halting progression of low-grade astrocytomas is lacking. No lesion should undergo radiotherapy without a tissue diagnosis. This modality may be primary treatment for (1) inoperable tumors and (2) aggressive lesions such as anaplastic astrocytomas and glioblastomas. Radiotherapy may be useful for (1) residual tumor after surgery and (2) recurrent tumor, but controversy exists. A dose of 50 Gy is delivered to the tumor in daily fractions of 1.5-2 Gy. This dose has not been shown to be curative in most studies. Some series report local failure rates reduce when more than 50 Gy is administered. Pitfalls include (1) acute and delayed myelopathy, (2) diminished skeletal growth in young children, and (3) increased difficulty with subsequent surgical tumor removal. This is particularly important if radiotherapy does not control the growth of the lesion. Chemotherapy This is considered experimental in the treatment of spinal cord tumors. A number of protocols are undergoing examination, primarily with childhood astrocytomas. Intraoperative details: The spinal cord is sensitive to decreased perfusion, and hypotension should be avoided. A surgical field clear of blood is necessary to minimize morbidity. For a tumor that spans several spinal levels, perform a wide laminectomy or laminoplasty. The tumor is localized visually with intraoperative ultrasound or spinal stereotaxy. Perform myelotomy at the thinnest area between the tumor and spinal cord. The myelotomy should be made in a linear fashion to spare vertically running white matter tracts. The incision generally is made in the midline, although occasionally eccentric lesions may be approached through the dorsal root entry zone. If the tumor has an exophytic component, this is the initial area of approach. The operating microscope provides magnification and illumination in this critical area. The Cavitron ultrasonic surgical aspirator (CUSA) or the carbon dioxide laser may be of value. Debulk any exophytic component prior to addressing a tumor located within the parenchyma. Tumors tend to be avascular and may have a true capsule or a definable plane between the tumor and cord, facilitating complete removal. Monitoring spinal cord function using intraoperative electrophysiology, such as somatosensory-evoked potentials and motor-evoked potentials (see Picture 9), may reassure the surgeon during the procedure and perhaps lead to improved outcomes. Total removal of the neoplasm, with preservation of neurologic function, is the goal in most cases. Dural grafting may be necessary. Watertight dural closure is necessary to minimize formation of a pseudomeningocele or CSF leak. Tissue adhesives placed over the suture line may be of value for obtaining a watertight closure. Postoperative details: Patients with cervical tumors should be considered for continued mechanical ventilation in the immediate postoperative period. CSF leakage should be treated aggressively with suture closure, collodium, lumbar drainage, or reoperation for closure. Steroid usage may inhibit wound healing and predispose to CSF leakage; hence, institute rapid tapering where feasible. New-onset urinary retention may require prolonged bladder catheterization, either continual or intermittent. A bowel stimulation regimen may be necessary for new abnormalities. Physical therapy and occupational therapy should be instituted early in the postoperative course. Follow-up care: Patients are followed clinically. Increasing symptoms or new neurological deficits should lead to a search for tumor growth in the tumor. MRI the day after surgery gives the best estimate of completeness of resection. Subsequent MRIs are performed to search for signs of recurrence or residual tumor growth. The optimal timing of follow-up imaging has not yet been determined. Residual tumor can be considered for repeat resection, radiation therapy, or observation. If tumor recurrence is noted, imaging the entire neuraxis is warranted because even benign ependymomas may change their growth characteristics and produce seeding. The majority of patients have an increased deficit during the immediate postoperative period. This may be due to edema from surgical manipulation or an alteration in blood flow. Most deficits return to preoperative baseline within 3-6 months. Complications are as follows:
Prognostic factors include the following: Tumor histology Histologically and clinically aggressive tumors have a poor prognosis despite treatment. Radical surgical removal can lead to severe neurologic impairment. Anaplastic astrocytoma Removal of tumor has not been shown to be of value, with survival of less than 2 years. Patients with low-grade astrocytomas and ependymomas may have prolonged survival with total excision. Patients with developmental tumors and lipomas may have prolonged survival despite residual tumor after surgery. Severity of preoperative neurologic deficit Neurologic morbidity is associated with preoperative functional status. Individuals with mild-to-moderate deficits often improve significantly following surgical removal, while those with advanced neurologic compromise generally have no worthwhile improvement. This emphasizes the need for early intervention and close follow-up. Age Advancing age (>60 y) is a negative prognostic factor. Completeness of resection Total removal of a benign tumor may result in long-term control or cure. Location of lesion Higher morbidity is associated with surgical removal of upper thoracic and conus lesions. Size of lesion Tumors spanning several levels may produce a corkscrew growth pattern that requires extensive dissection of the spinal cord in order to expose the tumor. Arachnoid scarring and cord atrophy These are negative prognostic factors for ependymomas. Syrinx The presence of a syrinx suggests a noninfiltrative lesion and carries a better prognosis.
Whereas the value of total excision of ependymomas is clear, the value of radical resection of astrocytomas is less certain. If an easily defined plane around the tumor can be followed and complete removal achieved, management is rather straightforward. However, if an ill-defined plane is present, the risk-to-benefit ratio for aggressive removal is not clear. The role of radiotherapy in the management of slowly growing tumors is not clear. Total excision of ependymomas does not warrant further treatment. This also probably is true of astrocytomas. However, the best management of residual or recurrent tumor is not defined. Whether repeat excision, watchful waiting, or radiation therapy is the best choice needs to be clarified. Intraoperative electrophysiologic monitoring is thought to be useful, but its efficacy is unproven. Although MRI greatly facilitates diagnosis of these lesions, pressure to control health care costs may delay diagnostic testing of mildly symptomatic patients. Currently, no satisfactory modality is available to affect the relentless course of malignant astrocytomas. Novel therapies need to be developed. Stereotaxic radiosurgery has found a place in the management of intracranial tumors. With anticipated future developments, spinal radiosurgery may have a role in management. Developmental tumors can be quite adherent to the spinal cord. The role of radical surgery to remove all traces of the tumor is not completely understood. Development of neuroprotective agents for use during surgery warrants further study. Management of these potentially debilitating and treacherous lesions has come a long way in the last 100 years. Advances in imaging and surgical technique have led to removal of many tumors, with high success and low morbidity. However, the relative rarity of the tumor, along with its slow growth characteristics, makes the accumulation of large patient series difficult. Presently, in many situations, the clinician can only care for patients harboring intramedullary spinal cord tumors using an incomplete knowledge base regarding the optimal management.
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
){kind=small}
