Spinal epidural metastasis: implications for spinal analgesia to treat "refractory" cancer pain.

Two hundred one consecutive patients with cancer pain who received intrathecal pain treatment between 1985 and 1993 were included in this retrospective study undertaken to test the hypothesis that epidural metastasis is a common cause of "refractory" cancer pain and that its presence may affect the efficacy and the complication rates of intraspinal pain treatment. Fifty-seven (approximately 28%) patients were investigated by metrizamide myelography, computerized tomography (CT), magnetic resonance imaging (MRI), laminectomy, or neurohistopathology. Epidural metastases were found in 40 (70%) and spinal stenosis in 33 (approximately 58%); 7 patients with total and 26 with partial occlusion of the spinal canal. Presence of epidural metastasis affected catheter insertion complications, daily dosages, and complications of the intrathecal pain treatment only when it was associated with spinal canal stenosis (partial or total). During the period of the intrathecal treatment, the patients with confirmed epidural metastasis and total spinal canal stenosis needed significantly (P < 0.05) higher daily doses of opioid (means = 77 +/- 103 versus 22 +/- 29 mg) and intrathecal bupivacaine (means = 65 +/- 44 versus 33 +/- 20 mg) and had significantly (P < 0.05) higher rates (14% versus 0%) of radicular pain at injection and poor distribution of analgesia than those without epidural metastasis and spinal canal stenosis. In contrast, the rate of occurrence of post-dural puncture headache was significantly (P < 0.05) lower in patients with partial (4%) and total (14%) spinal stenosis than in those without (29%). Unexpected paraplegia occurred in four patients and was due to accidental injury during attempted dural puncture (N = 1) and collapse (due to cerebrospinal fluid leakage leading to "medullary coning" of an unknown epidural metastasis (N = 3).