Pain

Pain, the most common first presenting symptom of spine metastasis, can have different mechanisms depending on the tumor’s interaction with the bony spine or the neural elements, and 83% to 95% of patients with spinal metastases complain of pain. Because pain often precedes other neurological symptoms, it should be carefully evaluated by clinicians.^5,6 If tumor compresses a nerve root, pain that is burning and radiates down the leg in a dermatomal distribution (radicular pain) is a common symptom. If tumor causes a pathological fracture or collapse of a vertebral body, that collapse can narrow the foramina and cause radicular pain, or it may cause mechanical pain. Mechanical pain typically is worse with loading of the spine during sitting or standing, and it often does not improve with anti-inflammatory medications. Local or biological pain may also be experienced by patients with spine metastases; it is usually described as a deep ache that is worse at night. This type of pain can improve with anti-inflammatory medications or corticosteroids, and it is likely due to inflammation in the spine due to the presence of tumor. Tenderness to palpation over the spine may be evident on physical examination due to local pain from periosteal stretching and inflammation. Treatment of spine metastases often includes a goal of treatment of the patient’s pain, and the type of pain a patient experiences may guide that treatment. Local pain is often palliated with radiation treatment; mechanical pain may be best addressed with bracing or surgical stabilization.

Neurological Dysfunction

The second most common group of symptoms that patients with spinal metastases complain of is neurological deficit.⁷ Weakness in upper or lower extremities may result from epidural compression of the spinal cord, individual nerve roots, or the cauda equina by tumor. Tumor or fractured bone fragments may impinge on neural structures, and patients may also complain of autonomic problems including bowel, bladder (usually urinary retention), or sexual dysfunction. Asking patients directly about these problems should be routine, as the symptoms might not be revealed on initial history gathering. Without treatment, motor dysfunction usually progresses to paralysis. Dermatomal changes in sensation, including anesthesia, hyperesthesia, and parasthesia, usually occur with motor dysfunction and pain. Patients with spinal cord compression may complain of sensory abnormalities in a band-like distribution across the chest or abdomen. Myelopathy also results in hyperreflexia from chronic spinal cord compression. Diagnosis before these neurological deficits occur is important because neurological prognosis is related to the amount of neurological function at the time of diagnosis.⁴ Pain is usually experienced before a neurological deficit, but because of the extremely common prevalence of back pain in the general population, metastatic spine disease can be missed until deficit occurs. Any patient with a known history of cancer and a new complaint of back or neck pain should be thoroughly evaluated for spinal metastatic disease. As the thoracic spine is the most common location for metastasis to the spine, and degenerative problems there are less common, pain in the thoracic spine should cue clinicians to consider a neoplastic process in patients with new-onset thoracic pain.

Imaging Studies

Plain radiographs often serve as an initial imaging evaluation of a patient with back pain, owing to their low cost and widespread availability. These studies may help in identifying significant abnormalities such as compression fractures, scoliosis, large lytic or sclerotic osseous lesions, and radiopaque extraossous lesions. However, in order to be apparent on plain radiographs, the lesions must reach a significant size, thereby making these studies a fairly insensitive modality in tumor diagnosis.⁸ Although radiographs may be an appropriate initial study in patients without history of cancer, they should not be used as a screening or diagnostic modality in patients with an elevated suspicion of spine tumors.

Multidetector computed tomography (CT) scanners provide a rapid high-resolution image of the spine and surrounding structures. The availability of two- and three-dimensional reconstructions allows easy localization and visualization of lesions that are hyper- or hypodense in relation to the surrounding structures. The lesions appear lytic, sclerotic, or mixed and this information may aid in the determination of the potential stability of the spine. The great definition of the osseous anatomy also aids in operative planning, allowing precise measurements of the screw length and diameter suitable for each vertebral body, as well as the trajectory of the screw insertion. CT imaging is also invaluable in imaging instrumentation position and the degree of osseous fusion. Finally, injection of dye into the subarachnoid space during CT myelography allows precise definition of the CSF surrounding the cord and nerve roots. This technique is invaluable in diagnosing neural element compression in the setting of instrumentation that creates ferromagnetic artifact on magnetic resonance imaging (MRI) or in patients who cannot undergo MRI. It has become a standard component of treatment planning for spine stereotactic radiosurgery in patients with spinal hardware.

Universal nuclear body scans have become a standard component of histological staging for certain cancers. Nuclear scintigraphy (bone scan) permits detection of osseous remodeling throughout the skeletal system and has been reported to have 62% to 89% sensitivity in detection of spinal metastases.⁹ However, it is not specific for neoplasms and cannot differentiate tumors from regions of infection and inflammation. Furthermore, in order for a neoplasm to be detected on a bone scan, active osseous remodeling must be taking place, which is not always the case in sclerotic neoplasms. Single-photon emission computed tomography (SPECT) permits improved differentiation of neoplastic from inflammatory processes, because it detects the metabolic components of the lesion.^10,11 It allows three-dimensional imaging of the lesion with sensitivity and specificity superior to standard bone scans. Finally, positron emission tomography (PET) using fluoride-18 (¹⁸F-PET) and ¹⁸F-fluorodeoxyglucose (¹⁸FDG-PET) provides a sensitive screening modality for neoplastic lesions through the body. ¹⁸F-PET detects regions of fluoride uptake and thereby skeletal remodeling. ¹⁸FDG-PET detects regions of high glucose uptake in the skeletal system and the soft tissues, thereby detecting hypermetabolic lesions that may signify a neoplastic, degenerative, inflammatory, or infectious process. However, correlation with concomitant MRI and CT scans generally allows high diagnostic specificity. Furthermore, ¹⁸FDG-PET has been shown to be a highly sensitive and specific screening modality in patients with solid tumors harboring spinal lytic and mixed lesions.¹² The decreased sensitivity of ¹⁸FDG-PET in detection of sclerotic lesions may be associated with the acellular and therefore low metabolic nature of these lesions. PET imaging may also aid in the selection of biopsy targets in the setting of multiple lesions, because lesions with higher metabolic activity will likely produce higher diagnostic yield due to their high cellularity.

MRI remains the gold standard in spine imaging and in detection of spinal metastases. MRI has superior sensitivity and specificity in screening for vertebral metastases, allowing precise delineation of their size, quantity, and location. It furthermore provides thorough information about the surrounding soft tissues that include the spinal cord and nerve roots, ligaments, meninges, interverterbal disks, and paraspinal musculature. In addition to the standard T1- and T2-weighted sequences with gadolinium injection, fat suppression and diffusion-weighted sequences may provide additional information about the acuity and etiology of visualized fractures. MRI allows excellent visualization of the spinal cord and nerve roots and any compression that may result from epidural tumor extension. T2 hyperintensity of the cord signifies cord edema and may indicate increased severity of compression.

Percutaneous CT-guided biopsy is a frequent step in the diagnostic algorithm, because 10% to 20% of the spine metastases have no known source and determination of tumor histological type is of paramount importance in determining the appropriate treatment.¹³ Furthermore, in patients with distant cancer history a biopsy may be required in order to rule out a second malignancy. Modern large-bore needle biopsy techniques have excellent diagnostic yield and are generally performed as an outpatient procedure. Caution must be exercised when interpreting negative biospsies of blastic lesions, because they may be falsely negative owing to their acellular nature. Such patients may merit additional follow-up in order to monitor these lesions clinically and radiographically.

Metastases that originate from hypervascular primary tumors generally require preoperative digital subtraction angiography and embolization. Angiography allows evaluation of the vascularity and the blood supply of the tumor and delineation of the location of the artery of Adamkiewitz. Renal, thyroid, hepatocellular, neuroendocrine tumors and tumors that contain “angio” or “hemangio” in their name should generally undergo angiography and embolization when possible. Complete and even partial preoperative embolization of vascular tumors has been shown to significantly decrease intraoperative blood loss.^14–16

Surgery

The most convincing data supporting surgical decompression resulted from a prospective randomized trial conducted by Patchell and associates.²³ The trial included patients with solid metastatic spinal tumors causing spinal cord compression, which was defined as any degree of cord displacement. The patients had to have at least one neurological sign or symptom, which may have been pain, and could not be paraplegic longer than 48 hours. The trial excluded patients with radiosensitive tumors such as hematological and germ cell malignancies and patients with previous radiation that precluded the study radiation dose. Patients were randomized to two trial groups: the first group received 30 Gy fractionated radiation therapy and the second group first underwent surgical decompression followed by radiation. The study was stopped at the midpoint of the enrollment owing to the clear superiority of the surgery group in the ability to ambulate, duration of ambulation, maintenance of continence, decreased use of opioids and corticosteroids, and extended survival. The limitations of the study include randomization of patients with instability to the radiation arm as well as the surgery arm, outdated spinal stability criteria, low benchmark for ambulation (only four steps), and lack of requirement for instrumented stabilization in the surgery arm.

Several grading systems have been devised in order to tailor the extent of surgical intervention to the expected survival of the patient. Tomita and colleagues proposed a 10-point scale that considered three prognostic factors: tumor histological type, the extent of visceral metastases, and the number of skeletal metastases.²⁴ They recommended that patients with slowly growing tumors (breast, thyroid) and moderately growing tumors (renal, uterus) and solitary spine metastases undergo en bloc tumor excision. Patients with rapidly growing tumors (lung, stomach) and patients with more extensive visceral and skeletal burden were recommended a spectrum of intralesional excision, palliative decompression, or supportive care. Tokuhashi and co-workers proposed a similar scale which in addition to the above-mentioned factors included the general performance status of the patient and the extent of neurological deficits.²⁵ They used these factors in order to estimate the expected survival and recommended that patients with longer than 1-year survival should undergo excisional surgery (goal of en bloc spondylectomy), and patients with expected survival between 6 months and 1 year should undergo palliative surgery (piecemeal cord decompression and stabilization).

The choice of surgical strategy and approach for spinal metastases remains predicated on the surgeon’s comfort and training. The vertebral body represents the most common site of spinal metastases. Thus, anterior approaches often provide the most direct access to these tumors. The transcervical approach provides direct access to the subaxial cervical spine and most spine surgeons are facile with this approach. The transoral and transmandibular approaches have been traditionally used in order to access the craniocervical junction and the atlantoaxial spine; however, recently transnasal and transcervical approaches have also been employed in order to access these regions. Manubriotomy, sternotomy, or the trapdoor approach may be required in order to provide ventral access to the upper thoracic spine (T1-T4). However, the great vessels and mediastinal contents generally complicate ventral access to the upper thoracic spine. The above-mentioned complex anterior approaches to the craniocervical junction and the upper thoracic areas are generally used in order to treat primary tumors, since the palliative nature of metastatic tumor treatment rarely justifies the potential morbidity of these approaches. A thoracotomy provides direct vertebral body access at the T5-L1 levels.^26,27 Access to T5-T7 usually requires sacrifice of the rib one level above the target vertebral level, and access to T8-L1 requires sacrifice of the rib two levels above. In order to visualize T11-L1 vertebral bodies, the diaphragm usually has to be reflected. The retroperitoneal approach provides ventral access to L2-L5 and sacral regions. Head and neck, thoracic, general, and vascular surgeons may be required to provide these ventral approaches, and are recommended if the patient had previous radiation, anterior surgery, or extensive tumor burden ventral to the spine.

Patients with limited lung capacity and previous neck, thoracic, or abdominal surgery or radiation may not be optimal candidates for anterior approach surgery. Posterior approaches provide direct access to the spinal cord and thereby allow direct decompression without extensive osseous work.^28,29 While the posterior approach is rarely employed in order to access cervical vertebral bodies, costotransversectomy and transpedicular approaches provide access to the thoracic and lumbar vertebral bodies. In the thoracic spine, nerve roots below T2 may be sacrificed in order to obtain posterolateral access to the anterior column. In the lumbar spine, the nerve roots must be preserved and may be retracted medially in order to provide ventral access.

Combined anterior and posterior approaches may be implemented in order to achieve circumferential cord decompression and spinal reconstruction. Although en bloc tumor resection may be achieved via a posterior approach in the lumbar spine, a combined anterior-posterior approach is generally required in the thoracic spine. Combined approaches are generally associated with higher surgical morbidity.³⁰

Vertebral body reconstruction may be accomplished with the help of Steinman pins with polymethylmethacrylate (PMMA), PMMA in a chest tube, mesh or expandable titanium cages, or polyetheretherketone (PEEK) cages.³¹ When a ventral approach is employed, an anterior plate spanning one level above and one level below the vertebrectomy defect is generally used in order to buttress the cage or cement reconstruction. In posterior decompression, posterolateral lateral mass or pedicle instrumentation is usually employed, spanning at least two levels above and below the tumor.