Quantifying Trial-to-trial Variability of Intraoperative Transcranial Motor Evoked Potentials

Quantifying Trial-to-trial Variability of Intraoperative Transcranial Motor Evoked Potentials
Hale, B., Samusyte, G., Giampiccolo, D., Koltzenburg, M.

Background

Intraoperative neurophysiological monitoring (IONM) of transcranial motor evoked potentials from muscles (mMEPs) and directly from the spinal cord (D-waves) are common spinal surgery modalities. Alarm criteria often rely on peak-to-peak amplitude reduction yet there is known inherent amplitude variability. This retrospective study quantifies baseline mMEP and D-wave amplitudes to better understand the biological variability and potential impact on intraoperative alarm criteria.

Methods

IONM spinal surgeries below T1 level with APB mMEPs and separate C2-T6 intramedullary lesion surgeries with D-waves were included. mMEPs and D-waves were elicited using supramaximal constant-voltage stimuli and recorded on Medtronic Eclipse system. Data are consecutive baseline amplitudes from individual cases over pre-incision (mMEPs) or pre-myelotomy (D-waves) periods of less than 5 minutes with patients under steady-state total intravenous anaesthesia.

Results

At least three mMEPs were recorded in 20 patients, 12 of whom had 1-3 additional mMEPs. The grouped average of each patient’s mean amplitude (±SEM) was not different to the average of each patient’s first mMEP (3.4±0.2 vs 3.3±0.3mV, p=0.152). Group trial-to-trial variability, expressed as Mean Consecutive Difference normalised to first mMEP amplitude, was relatively low (median of 20%), but varied markedly between individuals (range 4-89%). Recording >3 mMEPs did not affect these parameters. The Minimum Detectable Change, indicating true amplitude change beyond biological variability, was 91% of the first mMEP in these 20 patients.

At least three D-waves were recorded in 14 separate patients, 11 of whom had 1-7 additional D-waves. The grouped average of each patient’s mean amplitude was not different to the average of each patient’s first D-wave (17.8µV (IQR 22) vs 18µV (IQR 22), p=0.551). Group trial-to-trial variability, expressed as Mean Consecutive Difference normalised to first D-wave amplitude, was low (median of 6%), but varied between individuals from 2 to 20%. Recording >3 D-waves did not affect these parameters. Minimum Detectable Change, indicating true amplitude change beyond biological variability, was 22% of the first D-wave amplitude in these 14 patients.

Conclusion

Data suggest 50% amplitude reduction alarm threshold in intrinsic hand mMEPs for spinal IONM would increase false-positives. 80% or 100% thresholds at supramaximal mMEP stimulation may better account for biological variability and be a more appropriate alarm criterion. In contrast, our data suggest the 30-50% criteria for D-waves would not cause false positives. Combined with previous studies linking D-wave amplitude to clinical outcomes, this supports an established 30-50% amplitude reduction threshold as a reliable warning criterion for D-waves.

References

Deletis, V. (2020). ‘Intraoperative neurophysiology and methodologies used to monitor the functional integrity of the motor system’, Neurophysiology in neurosurgery, pp. 17-34.

Journée, H.L., Berends, H.I. and Kruyt, M.C., 2017. The percentage of amplitude decrease warning criteria for transcranial MEP monitoring. Journal of Clinical Neurophysiology, 34(1), pp.22-31.

Outputs

Poster presentation at European Congress of Clinical Neurophysiology, Central Hall Westminster, London, September 9-12, 2025