Validity of evoked potential as biomarker for predicting early neural function changes after thoracic spinal decompression surgery in patients with neurological deficits

Objective To evaluate the validity of intraoperative evoked potential (EP) including motor evoked potential (MEP) and somatosensory evoked potentials (SEP) as a biomarker for predicting neural function changes after thoracic spinal decompression (TSD) surgery. Method A consecutive series of 336 TSD surgeries were reviewed between 2010 and 2021 from four spine center. All patients with TSD were divided into 3 groups according to different intraoperative EP results: group 1, EP alerts; group 2, no obvious EP deterioration; group 3, EP improvement compared with baselines. The lower limb Japanese Orthopedic Association (JOA) scores (as well as early and long-term JOA recovery rate) were utilized to quantitatively assess pre- and postoperative neural function change. Results Among the 3 subgroups according to the different EP changes, the early JOA recovery rate (RR%) in the EP improvement group was significantly better than the other two groups (51.3 ± 58.6* vs. 27.5 ± 31.2 and 33.3 ± 43.1; p < 0.01) after 3-month follow-up. The mean MEP and SEP amplitude were from 116 ± 57 µV to 347 ± 71 µV ( p < 0.01) and from 1.86 ± 0.24 µV to 2.65 ± 0.29 µV ( p < 0.01) between spinal cord pre-decompression and post-decompression. Moreover, multivariate logistic regression analysis revealed that risk factors of EP improvement were duration of symptom ( p < 0.001, OR 10.9) and Preop. neurologic deficit degree ( p = 0.013, OR 7.46). Conclusion The intraoperative EP can predict postoperative neural function changes as a biomarker during TSD. Patient with EP improvement probably has better prognosis for early neural function recovery. The duration of symptom and preoperative neurologic deficit degree may be related to intraoperative EP improvement.