Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Primary tumors of the spine are far more rare than metastatic tumors.
Management of spinal tumors is multidisciplinary, requiring a variety of treatment modalities.
Treatment goals range from local disease control or cure to symptom palliation, prevention of neurological deterioration, and spinal stabilization.
In the case of primary malignant bone tumors, long-term local control, survival, and potential cure are generally dependent on en bloc tumor resection.
Metastatic disease is by far the most prevalent form of spinal neoplasia, with an incidence of up to 20,000 new cases per year. Primary tumors of the spine are far more rare and account for only 10% of spinal tumors. They can be either benign or malignant. The incidence of primary spinal neoplasms is estimated to be between 2.5 and 8.5 per 100,000 per year. Benign tumors are more common in young patients and involve the posterior elements, whereas malignant tumors tend to occur in the anterior column and affect elderly patients. Overall, primary spinal tumors are more common in men than in women, with osteoid osteoma, osteoblastoma, osteochondroma, plasmacytoma, chordoma, and chondrosarcoma more common in men than in women at nearly a 2:1 ratio.
The management of spinal tumors is multidisciplinary, requiring a team of diverse clinicians and a variety of treatment modalities, with goals ranging from local disease control or cure to symptom palliation, prevention of neurological deterioration, and spinal stabilization. The purpose of this chapter is to review the clinical presentation, diagnosis, and management of patients with primary tumors affecting the spine.
Spine tumors can present with varied symptomatology. The initial symptoms of a given tumor are often dictated by its location within the vertebra and its growth rate. Tumor growth itself can lead to periosteal irritation, resulting in a deep, unremitting, aching pain that is not related to movement, position, or effort. Classically, this pain is worse at night and is considered to be biological in nature. The destruction of bone or ligaments by tumor growth may cause spinal instability, which results in pain that is worse with movement and the upright position, improved with rest or when recumbent, and considered to be mechanical—not biological—in nature. Pathological fracture, progressive deformity, or tumor invasion may cause neural canal or foramen compromise, resulting in neurological symptoms from root or cord compression. These neurological symptoms may include radiculopathy, sensory deficit, weakness, or bowel/bladder incontinence. Epidural extension may also cause epidural venous congestion and hypertension, resulting in spinal cord edema, frank hemorrhage, demyelination, or ischemia. In truth, most tumors cause symptoms through a combination of the above mechanisms.
Neurological deficits may manifest themselves as frank lower extremity weakness if spinal cord compression occurs over a short period of time or as the result of a pathological fracture with bony compression. In cases where epidural compression occurs more gradually, neurological symptoms may develop more subtly as myelopathy with spasticity, ataxia, loss of sensation, and paraparesis progressing to paraplegia. Autonomic symptoms may also develop, most commonly as bowel and bladder incontinence, but also with signs of orthostatic hypotension and impotence.
The primary complaint of most patients with primary spine tumors is pain—in one review, more than 84% of patients complained of back pain (60.2%) or radicular pain (24%), with no apparent difference between benign or malignant histologies. In contrast, 55% of patients with malignant tumors were found to have neurological deficits, compared with only 35% of patients with benign tumors—a contrast that highlights the connection between the development of progressive neurological deficit and more malignant tumor growth. Admittedly, however, benign spinal neoplasms may also present with progressive neurological deficits.
Spinal tumors may also cause symptoms via paravertebral extension with involvement of surrounding structures, such as the paravertebral muscles or the ventrally located mediastinum and great vessels in the thoracic spine, or the vertebral arteries in the cervical spine.
The specific imaging modalities best suited to evaluate spinal tumors depend on the underlying pathology. As a guiding principle, any imaging should thoroughly evaluate all of the spinal segments involved. Generally speaking, the character of tumor borders is often dependent on rate of tumor growth. Clearly defined, sclerotic margins usually occur in slow-growing tumors, whereas faster-growing tumors lose their sclerotic rim and display mottled or ragged edges, with a surrounding pseudocapsule. Rapidly growing tumors also tend to be more destructive and display a lytic pattern of bone destruction. In general, tumor growth respects certain tissue planes, most notably the intervertebral disc, fascial planes, and spinal ligaments, although particularly malignant tumors may prove an exception to the rule.
Benign primary lesions tend to involve the posterior elements, whereas malignant lesions have a predilection for vertebral body and pedicles. Osteolysis and vertebral body collapse are common, but disc space height is usually maintained, even with advanced vertebral destruction.
Plain radiography can demonstrate the site of the lesion in question and, in some cases, can be diagnostic, although newer imaging modalities are quickly rendering plain films obsolete.
Radioisotope (technetium-99m) bone scanning is used to detect small osteoblastic lesions, although false positives may be seen with infection, fracture, or simple inflammation. This imaging modality is poorly suited for imaging rapidly growing lesions with a relative paucity of reactive bone formation, resulting in false negatives. Positron emission tomography (PET) scanning using fluorodeoxyglucose or sodium fluoride has become a highly sensitive tool for the identification and localization of tumors in the spine and in the body overall.
Computed tomography (CT) provides excellent resolution of bony anatomy and is excellent at capturing the extent of bony disease. Multiplanar image reconstruction aids in surgical staging and planning and, more recently, can even be used to facilitate stereotactic placement of spinal instrumentation. CT is limited, however, in its visualization of neural elements. CT myelography allows indirect visualization of the intradural contents, but it also exposes patients to procedural morbidity and requires unencumbered cerebrospinal fluid (CSF) flow in the caudal-cephalad direction, a process that may be limited by extensive epidural disease and obliteration of CSF channels.
Magnetic resonance imaging (MRI) is invaluable in identifying and characterizing tumors of the spine. Like CT, it provides multiplanar image reconstruction. Moreover, MRI yields unparalleled resolution of soft tissue and neural elements, often obviating the need for CT myelography. MRI is, however, limited by poor visualization/delineation of bony anatomy.
Other miscellaneous studies may have a role in the workup of spinal tumors, depending on the suspected pathology. Angiography is useful in delineating spinal vascular anatomy—such as the artery of Adamkiewicz—and tumor vascularity, which may lead to concomitant therapeutic embolization before surgical intervention. Such preoperative embolization can minimize intraoperative blood loss and postoperative morbidity, and should be considered in particularly bloody tumors such as aneurysmal bone cyst, giant cell tumors (GCTs), and hemangiomas. Thorough evaluation of the vascular extent of the lesion may provide valuable diagnostic information as well.
In any case where a spinal tumor is suspected, thorough imaging of the lesion in question is required. Additionally, however, a complete systemic workup should also be performed. CT of the chest, abdomen, and pelvis, or a PET scan is valuable in ruling out primary malignancies or other sites of disease. Ultimately, however, a tissue diagnosis is often required.
In cases where the diagnosis is unknown, obtaining tissue for histological diagnosis is the first step in the management of any spinal tumor. Biopsies must be planned with care as part of a comprehensive management strategy, to avoid tumor seeding along fascial planes or biopsy tracts; both increase risk of local recurrence. Areas of soft tissue extension or lytic destruction generally have the highest diagnostic yield. Some have suggested that resection of tissue along a biopsy tract can prevent recurrence caused by contamination of local tissues.
Biopsies can be separated into three categories: needle biopsy, open incisional biopsy, and open excisional biopsy. CT-guided or fluoroscopic-guided needle core biopsies allow for tissue diagnosis in up to 86% of cases and may be preferred over open biopsies, given their lower morbidity profile. The major limitation of needle biopsies is inadequate tissue sampling, which can result in a nondiagnostic specimen, as is often the case with densely blastic lesions, necrotic tumors, or vascular lesions.
If an open biopsy is selected, the biopsy incision should allow for an eventual excision of the biopsy tract during future definitive surgical approaches. Meticulous hemostasis must be ensured, as hematomas can track along fascial planes with tumor cells in tow. Bony margins must be conservative so as to avoid introducing instability or pathological fracture into a diseased spinal segment. Frozen sections must be sent intraoperatively to confirm adequate tissue sampling. Cultures should also be sent to rule out infectious pathology.
The goals of surgical resection—oncological cure, local control, or palliation—depend on the histology in question and, in the case of metastatic tumors, the burden of systemic disease. A number of authors have developed preoperative scoring systems to predict patient prognosis, and use said prognoses to guide general recommendations about the aggressiveness of treatment in the case of metastatic disease specifically. ,
Enneking described a staging system initially used for tumors of the long bone that was subsequently applied to primary tumors of the spine. The Enneking classification system helps to clarify goals of surgery and is a guide to adjuvant therapy. Both benign and malignant tumors can be classified using this system, and tumor grade is dictated by the local extent of disease and presence of metastasis.
Benign tumors are divided into three stages. Stage 1 refers to asymptomatic tumors surrounded by a true capsule; these tumors display an indolent growth pattern and rarely require surgery. Stage 2 refers to tumors growing actively that are symptomatic, surrounded by a thin true capsule, and adjacent pseudocapsule of reactive tissue; these tumors require intralesional or en bloc excision and recur infrequently. Stage 3 refers to tumors that are rapidly growing, with only a thin or incomplete capsule surrounded by a hypervascular pseudocapsule; these tumors are locally aggressive and require en bloc resection with intent for wide excision.
Enneking’s system divides malignant tumors into three stages as well, with each stage further subdivided into type A lesions, where tumor remains within the confines of the vertebra, and type B lesions, where tumor extends beyond the vertebra into adjacent soft tissues. Stage 1 lesions are low-grade, slow growing, and surrounded by a thick pseudocapsule that contains microscopic rests of tumor tissue. Stage 2 tumors are high-grade, rapidly growing lesions whose rapid growth precludes formation of a pseudocapsule and whose malignant growth pattern is characterized by recurrent seeding of the surrounding tissue with satellite nodules, skip metastases, pathological fractures, and epidural extension. Stage 3 lesions are similar to stage 2 tumors in their malignant growth pattern but are distinct in having metastasized to regional lymph nodes or distant organs. The Enneking grading system has been used to guide the aggressiveness of surgical resection, with resection strategies ranging from decompression and stabilization only to wide en bloc excision, or to palliative strategies for metastatic lesions.
Use of the Enneking system in the management of spinal tumors does, however, have some limitations, most of which center around the fact that the spine, unlike the appendicular skeleton, has critical adjacent structures that cannot be sacrificed without significant morbidity. The Enneking system does not account for the continuous nature of the epidural compartment, does not acknowledge the devastating implication of sacrificing the spinal cord and roots when wide local excision is recommended, and does not address the need to maintain spinal stability following aggressive oncological resection. Acknowledging these limitations, surgical interventions that are able to achieve the surgical margins recommended by the Enneking grading system might be deemed “Enneking Appropriate.”
The Weinstein–Boriani–Biagini (WBB) surgical staging system was designed to reflect the unique anatomic complexity of the spine and to dictate appropriate surgical approaches that spare the spinal cord without compromising surgical tumor margins. This system divides the vertebrae into 12 radiating zones, progressing in clockwise fashion from the spinous process (zone I) at 12 o’clock, as well as five concentric layers (A–E) in the transverse plane, with A representing extraosseous soft tissue, B superficial intraosseous, C deep intraosseous, D epidural tumor extension, and E intradural tumor spread. The cephalocaudal extent of the tumor is based on the number of spine segments involved. The system allows for a more reasonable approach to surgical planning by taking into account the underlying principle that en bloc excisions in the name of maximal oncological resection must be balanced by the need to preserve the spinal cord and neurological function. Depending on areas involved, the WBB system recommends anterior versus posterior approaches.
Chemotherapy, hormonal or immunological therapy, steroid administration, embolization, and external orthosis may play a role in the management of spine tumors to a varying degree. In the case of primary tumors, chemotherapy has traditionally played little role in the treatment of spinal tumors. More advanced chemotherapy regimens and the development of new targeted therapies have showed great promise, however, in the medical treatment of traditionally chemotherapy-resistant primary tumors.
Unfortunately, the terminology used in describing different types of surgical resections can be confusing. Surgical strategies can generally be divided into en bloc or intralesional resections. Curettage and intralesional resection generally refer to piecemeal removal of a tumor, whereas en bloc resection, or spondylectomy, indicates the intended removal of the whole tumor in one piece, with a layer of intact, surrounding healthy tissue ( ). The specimen is then submitted for histological study to confirm the actual extent of resection, regardless of the surgeon’s intent. A specimen will be labeled “intralesional” when the surgeon has cut within the tumor mass, “marginal” if the surgeon has dissected along a reactive pseudocapsule surrounding the tumor but without preserving a margin of normal tissue, and “wide” if the plane of surgical dissection is surrounded by a rim of continuous healthy tissue (>2 cm of healthy bone, reactive periosteum, or pleura). A wide resection is the intended goal of an en bloc procedure, but this goal is not always met. Intralesional resections provide symptom palliation but are less ideal from an oncological perspective, resulting in a high incidence of local recurrence because of the presumed seeding of residual tumor cells into the resection cavity.
The goal of surgical management is dictated by the pathology in question and the patient’s symptomatology. In general, however, surgery may or may not provide tissue diagnosis and should decompress neural elements, achieve spinal stability, and, if possible, cure the patient. Instances of acute neurological deficit, intractable pain, or progressive deformity usually require surgical intervention.
Video 46.1 En bloc spondylectomy.
For malignant primary tumors, long-term local control, survival, and potential cure are generally dependent on en bloc tumor resection. Marginal en bloc resection with regional adjuvant may be curative for aneurysmal bone cysts, GCTs, osteoid osteomas, and osteoblastomas, whereas chondrosarcoma, chordoma, Ewing sarcoma, and osteosarcoma require a true wide or marginal en bloc resection. In the case of chordoma, osteosarcoma, and chondrosarcoma specifically, achievement of an Enneking Appropriate resection has been correlated with lower local recurrence rates. Unfortunately, however, the extent of resection is often limited by the proximity of vital neural structures that cannot be sacrificed in an attempt to obtain surgical margins. However, some benign primary tumors do not require surgery. For example, hemangioma is often an incidental finding, and surgery may not be appropriate unless clinical signs and symptoms are present. As always, patient age, general wellbeing, and expected morbidity from surgical procedures must be considered.
Become a Clinical Tree membership for Full access and enjoy Unlimited articles
If you are a member. Log in here