What Are the Signs, Symptoms, and Treatments of Spinal Cord Compression?

Introduction and Scope of the Problem

Malignant spinal cord compression occurs in 2.5% to 5% of patients with cancer in the last 5 years of life. It presents as the initial manifestation of malignancy in up to 20% of new cancer diagnoses. The cumulative incidence varies by cancer type and is likely to continue to increase with improved cancer survival rates. Systemic cancers more likely to lead to spinal cord compression include prostate, breast, renal, and lung cancer; lymphoma; sarcoma; and multiple myeloma ( Table 60.1 ). The thoracic spine is the most common site for spinal cord metastases (70%), with the lumbar spine being second (20%); multiple levels are involved about one-third of the time.

Although in most instances spinal cord compression occurs in cancer as a result of metastatic spread to the spinal cord, it can also occur as a result of ischemic or hemorrhagic insults to the circulation involving the spinal cord, complications of radiation or chemotherapy, infections, or paraneoplastic syndromes. Occasionally, spinal cord compression occurs in the setting of primary spinal cord tumors such as astrocytomas and ependymomas. The focus of this chapter is metastatic spread to the spinal cord or malignant spinal cord compression (MSCC). The treatment of spinal cord compression is considered a medical emergency.

Historically, a diagnosis of MSCC was felt to portend an overall poor prognosis; however, it is known that patients with MSCC are a heterogenous population, with subsets of patients who share certain characteristics having improved survival. In a retrospective study of 176 patients with MSCC referred for surgical resection, median survival was 9.4 months after surgery. Primary cancer type was significant in predicting survival with lung, colorectal, bladder, esophageal, liver, and stomach cancers associated with a poorer prognosis and breast cancer, lymphoma, multiple myeloma, kidney, prostate, and thyroid cancers associated with a better prognosis. Other factors associated with decreased survival included older age, additional comorbidities, pathological spinal fractures, poor performance status, neurological deficits, and metastasis to the lung, liver, or brain.

In light of these findings, patient-based prognostic calculators have recently emerged in an effort to help with individual patient survival prediction. In the same retrospective study by Ahmed and colleagues, nine of these scoring systems were assessed for their accuracy in predicting short-term (30- and 90-day) and long-term (1-year) survival in patients by cancer type. Results from this analysis showed that the Skeletal Oncology Research Group nomogram developed by Paulino Pereira and colleagues incorporating different variables ranging from patient age to different lab values demonstrated the highest accuracy at predicting short-term survival. Tokuhashi and colleagues developed a prognostic scoring system that incorporates patient factors such as presence of metastatic disease and performance status to predict long-term survival. These tools can assist surgeons in a more personalized estimate of patient prognosis preoperatively.

Finally, in addition to better prognostication from improved understanding of the MSCC population, research has shown that treatment has increased the postoperative survival of patients with MSCC over the past 20 years. In a retrospective study of 1,515 patients who underwent surgery for MSCC from 1998 to 2017, patients with renal cancer had a 60% improvement in survival, patients with lung cancer had a 40% improvement in survival, and patients with colon cancer doubled their postoperative survival time.

Relevant Pathophysiology

MSCC usually spreads into the epidural space through extension of the tumor or through hematogenous spread via Batson’s venous plexus. Other, less common modes of spread into the spinal cord are through leptomeningeal and intramedullary spread. Leptomeningeal spread can take place through seeding of the meninges by the primary tumor (as can be seen in lung and breast cancer, melanoma, and lymphoma). Multiple myeloma typically presents with lytic bone lesions that extend into soft tissue as a result of activation of osteoclasts. Signs and symptoms are similar to those seen with epidural spread (see later discussion); however, lower-extremity weakness and paresthesias may be more pronounced. Intramedullary spread is relatively rare; in this instance, the primary tumor metastasizes into the spinal cord itself. Signs and symptoms are similar to those in other types of MSCC; however, weakness may be unilateral. Solid and hematological malignancies can also cause weakening of the bone leading to vertebral compression fractures. Bony fragments from a compression fracture may lead to compression of the spinal cord directly.

Summary of Evidence Regarding Diagnostic Workup

Whether to evaluate or confirm the diagnosis of MSCC is often the first diagnostic decision confronting a clinician. MSCC should always be considered early in a patient with cancer who presents with new back pain or new neurological deficits. Although imaging and treatment are often well tolerated, the burden of imaging in the setting of a frail patient with advanced cancer can be significant. MRI is the gold standard for diagnosis with a sensitivity of 93% and specificity of 97% but generally requires imaging time of well over 1 hour and necessitates sedation in some patients. The determination of whether to pursue confirmatory imaging should be made after clear dialogue with the patient and family. The decision should be based on the likelihood that the patient will benefit from treatment, that the burden of evaluation is acceptable, that significant pretest probability for MSCC exists, and that further evaluation and potential treatment are consistent with the patient’s goals. Pretest probability can be estimated with four factors on history and physical examination: an abnormal neurological examination, a known cancer diagnosis, a known vertebral metastasis at presentation, and new mid- to upper back pain. Patients with none of these four factors had a spinal cord compression incidence of 8% (notably not an insignificant percentage), and patients with three or four of these risk factors had an 81% likelihood of spinal cord compression.