Risk factors for malignant spinal cord compression and leptomeningeal metastases in cancer patients presenting to an urgent care center

Abstract

9606 Background: Malignant spinal cord compression is an oncologic emergency that is associated with considerable morbidity including irreversible loss of neurologic function. As many of the presenting signs and symptoms of spinal cord compression are common to other less-threatening conditions, it would be helpful to identify early clinical predictors specific for spinal cord compression. Methods: The medical records of all patients who presented to the emergency department of a cancer center and required spine MRI imaging in 2003 were reviewed. MRI findings were correlated with neurologic presentation. Results: There were 176 MRIs performed on 164 patients (84 men, 80 women; median age 63, range 22–96) with breast (n=31, 19%), prostate (n=29, 18%), lung (n=26, 16%), colon (n=9, 5%), myeloma (n=9, 5%), or other malignancies (60, 37%). Fifty two (29.6%) MRIs demonstrated spinal cord compression (SCC, defined as thecal sac compression with associated T2 signal change in cord) and 13 (7.4%) demonstrated incidental leptomeningeal metastases (LM). Multivariate logistical regression analysis identified four independent predictors of SCC: male gender (OR=2.95, 95% CI= 1.27–6.85, p=0.01), prior history of vertebral fracture (OR=9.09, 95% CI= 3.90–21.17, p< 0.0001), severe back pain (OR=4.21, 95% CI= 1.76–10.06, p= 0.001), and abnormal sensory exam (OR=9.45, 95%CI= 3.29–27.20, P<0.0001). SCC was noted in 13% of patients with only one risk factor and 90% of patients with 3 or more risk factors. Of note, 21 (40%) of scans demonstrating SCC were performed in patients without clinical evidence of weakness. The three independent risk factors associated with the presence of LM include: age less than 55 (OR=5.52, 95% CI= 1.40–21.7, p=0.02), extensor plantar reflex (OR=10.96, 95% CI= 2.77–43.46, p<0.001), and absence of tenderness upon palpation of spine (OR=8.02, 95% CI= 2.06- 31.23, p=0.003). Conclusions: Some of these findings are consistent with prior retrospective studies and provide a useful framework for assessing cancer patients in whom the diagnosis of SCC or LM are considered. In particular, absence of leg weakness does not exclude SCC, and the high rate of incidental LM demonstrates the importance of performing contrast enhanced scans. No significant financial relationships to disclose.