Several early studies have reported the effect of surgical decompression in cervical SCI. Kiwerski (1986) reported a retrospective series of individuals treated from the late 1960s to the time of publication. Individuals treated surgically and conservatively both had what was then an acceptable neurological outcome, but with surgery individuals showing superior results, with 66% and 49% of individuals showing neurological improvement at follow-up, respectively. Benzel et al. (1987) examined a series of surgically treated individuals with cervical SCI from 1979 to 1986 and demonstrated the possibility of neurological improvement with surgery even in groups of motor-complete injuries; however, there was no non-surgical control group in this study.
With respect to thoracolumbar SCI, Benzel et al. (1986a) described a series of 105 operative cases predominantly treated with anterior-only decompression (a formerly preferred approach). Despite no significant neurological improvement in those individuals with motor complete injuries, individuals with motor incomplete injuries went on to experience only minimal permanent neurological deficit. In this series, surgical efficacy could not be assessed due to absence of a comparator group. Hu et al. (1993) published a retrospective comparison of anterior-vs-posterior decompression in thoracolumbar SCI, finding no difference between approaches in terms of neurological outcome. Ultimately it is difficult to pinpoint the appropriateness of these studies today due to the vast improvements in surgical instrumentation in the intervening decades. Some centres perform endoscopic decompression in the setting of thoracolumbar SCI, with the thought of avoiding the significant morbidity of a thoracolumbar approach, while also achieving decompression of the compressive abnormality culprit (often the vertebral body). Beisse et al.(2005) report such a series, collected prospectively, that demonstrated comparable neurological outcomes to open surgery. Limitations to this approach include the inability to correct significant deformity.
Direct injury to the conus medullaris can present in a variety of neurological syndromes, as upper motor neuron fibres coalesce here and synapse with all anterior horn cells for the caudal lumbar and sacral spine. As such, a mix of upper motor neuron and lower motor neuron deficits to motor, bowel and bladder occur, depending on the specific anatomy of the injury. With respect to decompression of traumatic conus injuries, Rahimi-Movaghar et al. (2006) assessed a retrospective case series of individuals who primarily had suffered lumbar burst injuries with associated conus injuries. In this group, approximately 40% of individuals experienced neurological improvement after surgery, and a large majority experiencing an improvement attributed to recovery of an adjacent nerve root. As there was no non-operative control group, comparative efficacy of surgery could not be studied. The authors reported no effect of surgical timing on neurological outcome.
In addition to cord injury, there can often be an associated nerve root injury at or adjacent to the level of spinal trauma. In an important retrospective analysis, Benzel et al. (1986b) described substantial rates of improved nerve root function at the level adjacent to, or directly above, the injured spinal cord with surgical decompression of the neural foramen at the time of primary surgery. In contrast, none of the individuals who did not receive neural foramen decompression demonstrated nerve root function improvement. This suggests the reasonableness of pursuing nerve root decompression during surgery for SCI, despite nerve root compression not being a surgical indication in itself.
With respect to environments with economic limitations, Ojo et al. (2017) reported outcomes among a group of surgically treated individuals with SCI in Nigeria. Of 35 individuals, nine demonstrated improved Frankel grade at 6 months, and the group had an overall acceptable complication profile. Although there was no comparator group, the authors propose decompression and stabilization surgery as a reasonable proposition to enhance functional outcome, despite the challenges of complex spinal instrumentation in their economic setting.
There is level 3 evidence (based on 1 case control; (Kiwerski, 1986) that cervical decompression results in improved neurological functioning in comparison to conservative treatment. This is supported by level 4 evidence (based on 1 case series; (Benzel & Larson, 1987) showing neurological improvement but without a control group.
There is level 4 evidence (based on 1 case series; (Benzel & Larson, 1986a) that thoracolumbar decompression results in improved neurological functioning among individuals with incomplete SCI but not complete SCI.
There is level 4 evidence (based on 1 case series; (Hu et al., 1993) that there is no difference in neurological improvement after thoracolumbar decompression by either an anterior or posterior approach.
There is level 4 evidence (based on 1 pre-post study; (Beisse et al., 2005) that endoscopic thoracolumbar surgery results in improved neurologic outcomes for both motor complete and incomplete SCI, although those with incomplete injuries have greater rates of improvement.
There is level 4 evidence (based on 1 case series; (Rahimi-Movaghar et al., 2006) that decompression of lumbar burst injuries with associated conus injuries results in neurological improvement and recovery of adjacent nerve root.
Cervical decompression may improve neurological functioning post SCI.
Thoracolumbar decompression may improve neurological functioning among those with incomplete, but not complete SCI. Anterior and posterior approaches may be equally effective. Endoscopic approaches may be similarly effective to open decompression approaches.
Decompression surgery for lumbar burst and conus injuries may improve neurological outcomes and adjacent nerve root.