Immediate spinal cord decompression for cervical spinal cord injury: feasibility and outcome

The effect of immediate surgical spinal cord decompression on neurologic outcome after spinal cord injury is controversial. Experimental models strongly suggest a beneficial effect of early decompression but there is little supportive clinical evidence. This study is designed to evaluate the feasibility and outcome of an immediate spinal cord decompression treatment protocol for cervical spinal cord injury in a tertiary treatment center. To address this issue, 91 consecutive patients with acute, traumatic cervical spinal cord injury (1990-1997) were prospectively studied. Sixty-six patients (protocol group) underwent emergency magnetic resonance imaging (MRI) to determine the presence of persistent spinal cord compression followed, if indicated, by immediate operative decompression and stabilization. Twenty-five patients were managed outside the treatment protocol because of contraindication to magnetic resonance imaging, need for other emergency surgical procedures, or admitting surgeon preference (reference group). The protocol and reference groups had similar sex and age distributions, admitting Frankel grades, levels of neurologic injury, and Injury Severity Scores. Twenty-seven percent of patients seen were not enrolled in the treatment protocol because of the need for other emergent surgical treatment, contraindication to MRI, and specific surgeon bias regarding the “futility” of emergent treatment. The neurologic outcome for the patients in the reference group were similar to that in the previously reported literature. Fifty percent of protocol patients, compared with only 24% of reference patients, improved from their admitting Frankel grade. Eight protocol patients (12%), but no reference patients, improved from complete motor quadriplegia (Frankel grade A or B) to independent ambulation (Frankel grade D or E). Protocol patients required shorter intensive care unit stays, and shorter total hospital stays than reference patients. In the treatment protocol group, spinal cord decompression, confirmed by MRI, was achieved with immediate spinal column alignment and skeletal traction in 32 patients (46%). Thirty-four patients (54%) required emergent operative spinal cord decompression because of MRI-documented persistent spinal cord compression. We conclude that immediate spinal column stabilization and spinal cord decompression, based on magnetic resonance imaging, may significantly improve neurologic outcome. The feasibility of such a treatment protocol in a tertiary treatment center is well demonstrated. Additional multicenter trials are necessary to achieve definitive conclusions regarding clinical efficacy.