Several key areas must be considered in the diagnosis and management of spinal cord compression. Because the outcome can be devastating, a diagnosis must be made early and treatment initiated promptly. Although any malignancy can metastasize to the spine, clinicians should be aware that this occurs more commonly in certain diseases, ie, lung cancer, breast cancer, prostate cancer, and myeloma. The current algorithm for early diagnosis of spinal cord compression involves neurologic assessment and magnetic resonance imaging of the entire spine. Treatment generally consists of intravenous dexamethasone followed by oral dosing. Depending on the extent of the metastases, symptoms may also be managed with nonnarcotic pain medicines, anti-inflammatory medications, and/or bisphosphonates, with local radiation administered as needed. Surgery has often led to destabilization of the spine.
Much of the discussion surrounding the problem of spinal cord compression is based on perceptions about the problem that are inaccurate. For example: 1. Perception: Actual compression of the spinal cord is common. In fact, actual compression of the cord is uncommon. Impingement on the epidural space with significant local pain is, however, a common accompaniment of advanced cancer. 2. Perception: Surgery is the primary therapy for spinal cord compression. Rarely, if ever, is surgery the first option for treating spinal cord compression. Only when there are clear-cut signs of myelopathy, the structural integrity of the spine is compromised, the involved area of the spine has had full irradiation previously, or the diagnosis cannot otherwise be made, should surgery be employed initially. Posterior laminectomy to decompress the spinal cord is virtually never indicated. 3. Perception: An orderly set of tests starting with plain films of the spine and escalating to bone scintigraphy, computed tomography (CT) myelogram, or magnetic resonance imaging (MRI) is the best means by which to diagnose spinal cord compression. As soon as a cancer patient develops new back pain or worsening in a prior area of pain, he or she should move directly to a T1-weighted sagittal MRI of the entire spine. This is the quickest, most accurate, and most cost-effective means of diagnosing spinal cord compression early. 4. Perception: A good history and neurologic exam can determine which patients need further testing. There is no constellation of symptoms or signs that safely predicts which patient has significant epidural compression with a high potential for catastrophic loss of neurologic function. The simple presence of new back pain should initiate the work-up. There are several key areas that must be considered in the diagnosis and management of spinal cord compression. Given that the outcome (paralysis below the level of compression and the loss of bowel and bladder function) is so devastating, it is critical to diagnose the problem early. Once diagnosed, it is critical to move promptly to the proper therapy. Finally, it is important to follow these patients closely for the development of later cord compressions as they attain longer survival with systemic therapy. Early Detection of Spinal Cord Compression Historically, the issue of early detection of spinal cord compression was always discussed from the wrong end of the problem.[1-3] All of the early reports started with patients who had developed significant neurologic deficits, often with complete paralysis, and attempted to work backwards to a set of signs or symptoms that could be used to diagnose the problem earlier. This resulted in a set of recommendations that were centered on a complete neurologic examination with testing for subtle signs of myelopathy (Table 1). This "state of the art," which persisted into the 1980s, led to most patients being diagnosed with neurologic dysfunction of varying degrees and a significant number not recovering function.[1-3] Rodichok and colleagues at Albany presented the first successful algorithm for early diagnosis of spinal cord compression in 1981. They reasoned that the earlier trials had all shown that back pain preceded the onset of neurologic dysfunction by weeks to months and that "new onset back pain in a cancer patient" should be the trigger for starting the diagnostic work-up rather than "early signs of myelopathy." They demonstrated that nearly 60% of cancer patients who had new back pain and a normal neurologic examination had radiographically demonstrable spinal metastases with compression of the epidural space by tumor. When the disease was found by using new back pain and an abnormal plain film of the spine, the extent of compression was less than that seen in patients with presenting neurologic deficits.[4,5] These investigators started with plain films of the spine and performed myelography (MRI was not available at that time) in all patients with evidence of bone destruction. Previously, "early signs of myelopathy on neurologic exam" was the trigger for myelography. With this new diagnostic algorithm based on positive plain films, fully 95% of patients remained ambulatory when treatment consisted of radiation alone, compared to the significant percentage of patients who entered the process ambulatory and left nonambulatory using earlier approaches. These results have now been confirmed in several reports.[ 6-10] However, even some contemporary reports often reflect a high proportion of patients presenting with evidence of neurologic dysfunction, ranging from mild changes in deep tendon reflexes to frank paraplegia and autonomic dysfunction.[11-14] Unless the patient delays coming for evaluation, none of these findings should be allowed to develop. Distinguishing Causes of Back Pain
It is critical that the clinician be constantly aware of this problem in patients with diseases that commonly metastasize to the spine-primarily lung cancer, breast cancer, prostate cancer, and myeloma, although this can occur with any malignancy. Multiple studies have demonstrated that there is no constellation of symptoms that can separate patients with "significant" back pain from those with benign causes of pain.[5,6,8-10] Patients often report that they have had back pain for years, which is consistent with the widespread prevalence of this symptom in the general population. However, on close questioning, they often admit to subtle changes in symptoms or location that are the first sign of impending cord compression. In addition, like healthy individuals, cancer patients often overexert themselves. The clinician must maintain a high degree of suspicion for any new pain that does not resolve promptly with the usual measures of rest and antiinflammatory treatment. On occasion, patients present with worsening neurologic symptoms in the lower extremity in the absence of any back pain whatever. This should suggest one of several problems: (1) brain metastases, (2) neoplastic meningitis,[ 17] (3) a paraspinal mass with invasion up the neural foramen (4) plexopathy,[ 18] or (5) an intramedullary metastasis. Concurrent spinal and brain metastases are a common clinical problem, particularly in patients with lung cancer. Consequently, an MRI of the brain is needed whenever the origin of the symptoms is not clear. Neoplastic meningitis more commonly presents with nonanatomic pain and neurologic dysfunction and re- quires a spinal tap for diagnosis. Paraspinal masses are classically described as invading the neural foramen in pediatric neuroblastomas, but the common adult cancers that do this are small-cell lung cancer and lymphoma, both of which are readily identified on CT imaging. Since the spinal cord ends at about the 12th thoracic vertebra, compressive disease originating in the lumbar spine will impact the cauda equina rather than the spinal cord. This should be apparent on neurologic exam or picked up on MRI prior to the development of symptoms. Intramedullary metastases are rare but are picked up on the MRI.[20,21] They often present as ataxia. Current Algorithm
With the advent of MRI, it became clear that diagnostic accuracy was significantly improved, and MRI virtually completely replaced myelography as the definitive diagnostic procedure.[ 22-25] Although some have continued to push for CT-based myelography,[ 6,7] the literature supports using this technique only in patients who cannot have an MRI for technical reasons (eg, pacemaker) or who are severely claustrophobic. In the mid-1990s, Ruckdeschel observed that most patients eventually required an MRI and that obtaining a plain film or bone scan seemed to only be delaying the process. At that time, a full MRI of the spine took several hours and several thousand dollars for the individual cervical, thoracic, and lumbar studies. Using a T1-weighted sagittal-only view (scanning MRI), the entire spine can be imaged in less than 1 hour and detailed views taken with T1 and T2 weighting and contrast for any "positive" areas.[26-29] Ruckdeschel went on to demonstrate that employing the scanning MRI as the first diagnostic test was the most rapid and cost-effective means of diagnosing spinal cord compression. To obtain these estimates, he took the original 140 patients from the Rodichok study[4,5] and compared the original diagnostic sequence (plain film to bone scan to MRI) to simply proceeding to the T1-weighted MRI directly (Figure 1, Table 2). Our current algorithm is illustrated in Figure 2. This approach for early diagnosis of spinal cord compression is not, however, relevant for the average emergency room patient with back pain who has no prior or new evidence of cancer. Rodichok et al also demonstrated that 5% to 10% of patients had unsuspected sites of disease elsewhere in the spinal axis, necessitating full spinal evaluation. Consequently, ordering a limited MRI of the lumbar or thoracic area will overlook other significant sites of disease, a problem that is resolved by using the full sagittal study described above. Treatment of Spinal Metastases With or Without Cord Compression Getting the Evaluation Underway
One must be cognizant of the fact that patients with cancer and new onset back pain do not always present at the most convenient times for diagnostic studies and often present in distant emergency rooms where the understanding of back pain in the cancer patient is very different. Because the Rodichok study showed that 10% of patients with back pain, abnormal plain films, and a normal neurologic examination had complete obstruction on myelography (confirmed by Schiff et al), I have always been a proponent of initiating the diagnostic work-up at the first possible moment.[ 23] However, that is often not feasible for reasons of transportation, staff or equipment availability, or patient preference. In addition, patients often confuse their symptoms with prior episodes of back pain or choose to downplay the symptoms because they assume the pain may be a sign of progressive cancer. How then does the busy clinician sort out which patients need immediate attention, no matter the social problems, and which can safely wait until the next workday? The history and neurologic examination are the best guide. Any patient with lower-extremity weakness, bowel or bladder incontinence, clonus, down-going toes on the Babinski exam, or whose sphincter is patulous must be evaluated and treated immediately. My practice is to initiate high-dose dexamethasone (100 mg IV)[31-33] and obtain an immediate T1-weighted sagittal MRI of the spine with full T1 and T2 studies, with and without gadolinium enhancement, for any questionable areas. When pressed by administration as to whether they need to call in a technician or whether the exam can wait, I confirm that this test must be done without delay. Given the results presented by Patchell et al at the 2003 meeting of the American Society of Clinical Oncology-which suggested a benefit for adding surgery in the setting of existing myelopathy-I also notify the neurosurgeon and the radiation oncologist so that we may all view the MRI simultaneously with the neuroradiologist. This is not a time to wait for a written report to be generated before acting. Patients complaining of severe new-onset back pain and/or radicular symptoms (pain running down an extremity) are at high risk of having cord compression, but in the absence of any signs of early myelopathy, I start them on a dose of intravenous dexamethasone (10-40 mg) and then oral dexamethasone at 4 mg every 6 hours if the MRI is unavoidably delayed. It is important to note that radicular pain in the thoracic area is described as pain radiating around from back to front in a band-like distribution. This may be mistaken for epigastric or pleural pain. In cases where there has been extensive destruction of the vertebral body, the radicular symptoms may be bilateral, again leading to confusion with other gastrointestinal or pleural disorders. Patients with new-onset back pain, or a change in previous back pain symptoms, but who have no evidence of myelopathy or radiculopathy can safely wait until the next day for an MRI, with the caution that they should call immediately if symptoms emerge. If the MRI will be delayed for several days, I may start them on low-dose dexamethasone. Spinal Metastases and No Cord Compression
When the MRI demonstrates the presence of vertebral metastases but no evidence of epidural space invasion or cord compression, any steroids that were started can be discontinued and a decision on therapy can be based on the patient's overall situation. If systemic therapy is imperative and the symptoms can be managed with nonnarcotic pain medicines, antiinflammatory medications, and/or bisphosphonates, then local radiation therapy may not be needed. Care must be taken to follow these areas closely, as disease may progress later. A change in prior back pain will usually herald a problem that needs full reevaluation. Alternatively, a course of radiation therapy may be given to the local area, usually with coverage two vertebrae above and below the lesion. Epidural Space Invasion But No Myelopathy
When the MRI demonstrates invasion of the epidural space but there is no evidence of neurologic dysfunction, the optimal course of treatment is local radiation. In the Rodichok series, over 95% of such patients remained fully ambulatory with radiation alone.[4,5] Depending on the extent of invasion, I usually initiate low-dose dexamethasone and ask the radiation oncologist to begin the taper during treatment. Progression of symptoms during the initial steroid therapy and irradiation or recrudescence of the symptoms during tapering are potential signs of an unstable clinical situation warranting urgent reanalysis. There are no trials pointing to an appropriate dose of radiation, but doses of 30 to 50 Gy over 2 to 4 weeks are common. When radiculopathy is present, the course is the same, but the steroids are maintained for a longer period before tapering and the index of suspicion for progression during radiation is higher. Signs or Symptoms of Myelopathy
Until recently, patients with myelopathy were treated in much the same way as described above, and surgery was reserved for signs of progression or later recurrence.[35-40] In 2003, Patchell and colleagues reported on a series of patients with myelopathy due to metastatic cord compression, demonstrating that those randomized to surgery plus radiation did better than those receiving radiation therapy alone. Many centers have now adopted this approach, although the underlying status of the patient's disease must always be taken into account. Surgical intervention in this disorder has had a checkered history. For many years, the standard of therapy was a posterior laminectomy to "decompress" the spinal canal. While the laminectomy often improved symptoms in the short term, it frequently led to further destabilization of the spine. Most (85%) of spinal metastases are to the anterior vertebral body. When this is extensively destroyed, the removal of the posterior elements only serves to weaken the spine further, and any palliation is very short-lived. Recurrent or Progressive Disease
With earlier diagnosis of spinal metastases now more common, an increasing number of patients survive long enough to develop a recurrence in the previously irradiated site, which usually precludes further radiation therapy. Anterior decompression of the spine with reconstruction has become a popular and successful means of dealing with this problem, and there have been several reports of its successful application in this setting.[ 38-40] This strategy may also be appropriate for patients whose spine is structurally unstable due to the extent of vertebral invasion. Anterior decompression and reconstruction is, however, a major procedure. It often requires the presence of thoracic, orthopedic, or gastrointestinal surgeons, depending on the location of the lesion, and is not appropriate for patients who will likely have only a very short survival. We reserve the procedure for patients with an expected survival of over 6 months. Conclusions The overwhelming problem facing us as clinicians is the early detection of spinal epidural metastases, not the management of patients with significant neurologic deficits. If the patient delays seeking help, we can only respond so quickly. But if we fail to heed a patient's complaints of back pain and he or she goes on to develop neurologic deficits, then shame on us as physicians. There are few studies in this arena, but most of the algorithms for successful diagnosis and management have been worked out.
The authors have no significant financial interest or other relationship with the manufacturers of any products or providers of any service mentioned in this article.
1. Gilbert RW, Kim JH, Posner JB: Epidural spinal cord compression from metastatic tumor: Diagnosis and treatment. Ann Neurol 3:40-51, 1978.
2. Greenberg HS, Kim J-H, Posner JB: Epidural spinal cord compression from metastatic tumor: Results with a new treatment protocol. Ann Neurol 8:361-366, 1980.
3. Posner JB: Neurological complications of systemic cancer. Med Clin North Am 63:783- 800, 1979.
4. Rodichok LD, Harper GR, Ruckdeschel JC, et al: Early diagnosis of spinal epidural metastases. Am J Med 70:1181-1188, 1981.
5. Rodichok LD, Ruckdeschel JC, Harper GR, et al: Early detection and treatment of spinal epidural metastases: The role of myelography. Ann Neurol 20:696-702, 1986.
6. Choucair AK: Myelopathies in the cancer patient: Incidence, presentation, diagnosis and management. I. Oncology 5:71-80, 1991.
7. Choucair AK: Myelopathies in the cancer patient: Incidence, presentation, diagnosis and management. II. Oncology 5:25-31, 1991.
8. Grant R, Papadopoulos SM, Greenberg HS: Metastatic epidural spinal cord compression. Neurol Clin 9:825-841, 1991.
9. Posner JB: Back pain and epidural spinal cord compression. Med Clin North Am 71:185- 205, 1987.
10. Byrne TN: Spinal cord compression from epidural metastases. N Engl J Med 327:614-619, 1992.
11. Helweg-Larsen S: Clinical outcome in metastatic spinal cord compression. A prospective study of 153 patients. Acta Neurol Scand 94:269-275, 1996.
12. Milross CG, Davies MA, Fisher R, et al: The efficacy of treatment for malignant spinal cord compression. Australas Radiol 41:137-142, 1997.
13. Hill ME, Richards MA, Gregory WM, et al: Spinal cord compression in breast cancer: A review of 70 cases. Br J Cancer 68:969-973, 1993.
14. Rades D, Heidenreich F, Karstens JH: Final results of a prospective study of the prognostic value of the time to develop motor deficits before irradiation in metastatic spinal cord compression. Int J Rad Oncol Biol Phys 53:975- 979, 2002.
15. Bonica JJ: Historical, socioeconomic and diagnostic aspects of the problem. I. The nature of the problem, in Carron H, McLaughlin RE (eds): Management of Low Back Pain, p 1. Bristol, Stonebridge Press, 1982.
16. Posner JB: Clinical manifestations of brain metastases, in Weiss L, Gilbert H, Posner JB (eds): Brain Metastases, p 196. Boston, GK Hall, 1980.
17. Jayson GC, Howell A: Carcinomatous meningitis in solid tumors. Ann Oncol 7:773- 786, 1996.
18. Gilbert MR, Grossman SA: The incidence and nature of neurologic problems in patients with solid tumors. Am J Med 81:951-954, 1986.
19. Punt J, Pritchard J, Pincott JR, et al: Neuroblastoma: A review of 21 cases presenting with spinal cord compression. Cancer 45:3095- 3101, 1980.
20. Sze G, Abramson A, Krol G, et al: Gadolinium- DTPA in the evaluation of intradural extramedullary spinal disease. Am J Neuroradiol 9:153, 1988.
21. Sze G, Krol G, Zimmerman RD, et al: Intramedullary disease of the spine: Diagnosis using gadolinium-DTPA-enhanced MR imaging. AJR Am J Roentgenol 151:1193-1204, 1988.
22. Li KC, Poon PY: Sensitivity and specificity of MRI in detecting malignant spinal cord compression and in distinguishing malignant from benign compression fractures of vertebrae. Magn Reson Imaging 6:547-556, 1988.
23. Ruckdeschel JC: Spinal cord compression, in Abeloff M, Armitage J, Lichter A, et al (eds): Clinical Oncology, pp 619. New York, Churchill Livingstone, 1995.
24. Carmody RF, Yang PJ, Seely GW, et al: Spinal cord compression due to metastatic disease: Diagnosis with MR imaging versus myelography. Radiology 173:225-229, 1989.
25. Smoker WR, Godersky JC, Knutzon RK, et al: The role of MR imaging in evaluating metastatic spinal disease. AJR Am J Roentgenol 149:1241-1248, 1987.
26. Ruckdeschel JC: Rapid, cost-effective diagnosis of spinal cord compression due to cancer. Cancer Control 2:320-323, 1995.
27. Mehta RC, Marks MP, Hinks, RS, et al: MR evaluation of vertebral metastases: T1- weighted, short-inversion time inversion recovery, fast spin-echo, and inversion-recovery fast spin-echo sequences. Am J Neuroradiol 16:281- 288, 1995.
28. Dwyer AJ, Frank JA, Sank VJ, et al: Short T1 inversion-recovery pulse sequence: Analysis and initial experience in cancer imaging. Radiology 168:827-836, 1988.
29. Stimac GK, Porter BA, Olson DO, et al: Gadolinium-DTPA- enhanced MR imaging of spinal neoplasms: Preliminary investigation and comparison with unenhanced spin-echo and STIR sequences. Am J Roentgenol 151:1185- 1192, 1988.
30. Schiff D, O'Neill BP, Wang C, et al: Neuroimaging and treatment implications of patients with multiple epidural spinal metastases. Cancer 83:1593-1601, 1998.
31. Weissman DE: Glucocorticoid treatment for brain metastases and epidural spinal cord compression: A review. J Clin Oncol 6:543-551, 1988.
32. Delattre JY, Arbit E, Thaler HT, et al: A dose-response study of dexamethasone in a model of spinal cord compression caused by epidural tumor. J Neurosurg 70:920-925, 1989.
33. Vecht CJ, Haaxma-Reiche H, van Putten WL, et al: Initial bolus of conventional versus high dose dexamethasone in metastatic spinal cord compression. Neurology 39:1255-1257, 1989.
34. Patchell R, Tibbs PA, Regine WF, et al: A randomized trial of direct decompressive surgical resection in the treatment of spinal cord compression caused by metastasis (abstract 2). Proc Am Soc Clin Oncol 22:1, 2003.
35. Young RF, Post EM, King GA: Treatment of spinal epidural metastases: Randomized prospective comparison of laminectomy and radiotherapy. J Neurosurg 53:741-748, 1980.
36. Sundaresan N, DiGiancinto GV, Krol G, et al: Spondylectomy for malignant tumors of the spine. J Clin Oncol 7:1485-1491, 1989.
37. Harrington KD: Anterior cord decompression and spinal stabilization for patients with metastatic lesions of the spine. J Neurosurg 61:107-117, 1984.
38. Siegal T, Siegal T: Surgical decompression of anterior and posterior malignant epidural tumors compressing the spinal cord: A prospective study. Neurosurgery 17:424-432, 1985.
39. Harrington KD: Anterior decompression and stabilization of the spine as a treatment for vertebral body collapse and spinal cord compression from metastatic malignancy. Clin Orthop 233:177-197, 1988.
40. Perrin RG, McBroom RJ: Anterior versus posterior decompression symptomatic spinal metastases. Can J Neurol Sci 14:75-80, 1987.