Alert Timing and Corresponding Intervention With Intraoperative Spinal Cord Monitoring for High-Risk Spinal Surgery.

STUDY DESIGN

Prospective multicenter study.

OBJECTIVE

To analyze the incidence of intraoperative spinal neuromonitoring (IONM) alerts and neurological complications, as well as to determine which interventions are most effective at preventing postoperative neurological complications following IONM alerts in high risk spinal surgeries.

SUMMARY OF BACKGROUND DATA

IONM may play a role in identifying and preventing neural damage; however, few studies have clarified the outcomes of intervention after IONM alerts.

METHODS

We analyzed 2867 patients who underwent surgery for high risk spinal pathology using transcranial electrical motor-evoked potentials from 2010 to 2016. The high-risk spinal surgery cases consisted of 1009 spinal deformity cases, 622 cervical ossification of posterior longitudinal ligament (OPLL) cases, 249 thoracic-OPLL cases, 771 extramedullary spinal cord tumor cases, and 216 intramedullary spinal cord tumor (IMSCT) cases. We set a 70% amplitude reduction as the alarm threshold for transcranial electrical motor-evoked potentials and analyzed the outcomes of the interventions following monitoring alerts and postoperative neurological deficits.

RESULTS

The true positive, false positive, true negative, false negative, and rescue cases of IONM comprised 126, 234, 2362, 9, and 136 cases, respectively. Most alerts and interventions occurred during correction and release in deformity cases, posterior decompression and dekyphosis in OPLL cases, and tumor resection and surgery suspension with steroid injection in spinal cord tumor cases; however, individual interventions varied. The rescue rates (number of patients rescued with intervention after IONM alert/number of true positive cases plus rescue cases) for deformity, cervical-OPLL, thoracic--OPLL, extramedullary spinal cord tumor, and IMSCT cases were 61.4% (35/57), 82.1% (32/39), 40% (20/50), 52.5% (31/59), and 31.6% (18/57), respectively.

CONCLUSION

Our prospective multicenter study identified potential neural damage in 9.5% of cases and 52% rescue cases using IONM. Although the rescue ratios for t-OPLL and IMSCT were relatively low, appropriate intervention immediately after an IONM alert may prevent neural damage even in high-risk spinal surgeries.

LEVEL OF EVIDENCE