Friday, September 28, 2018 1:00 PM–2:30 PM abstracts: a new look at imaging: 211. The feasibility of transcutaneous transabdominal stimulation of the lumbosacral nerve roots in lumbar spine surgery

Abstract

BACKGROUND CONTEXT Intraoperative neuromonitoring during lumbar spine surgery has primarily relied on triggered and/or free-run electromyography (EMG) in mapping neural structures in the surgical field. Transcranial motor evoked potentials (MEPs) provide monitoring of the spinal cord, with nonspecific myotome activation following. Their application in lumbar procedures is increasingly being considered and utilized for cauda equina and nerve root monitoring. MEPs are more sensitive to detect a variety of spinal cord injury mechanisms, including ischemia, and, as such, have a higher sensitivity and specificity than stimulated electromyography (sEMG) with a higher postoperative predictive value for spinal cord deficits. The drawbacks of MEPs include the requirement for total intravenous anesthesia (TIVA), difficulty in achieving complete and reproducible baseline stimulation responses in relevant myotomes, lack of agreement on alarm criteria, and the potential for MEP-related injuries like tongue lacerations. Most importantly, MEPs are not validated in isolated lumbar applications, where the cauda equina and nerve roots, rather than the spinal cord, are at risk. There have, however, been several studies that have evaluated the utility of MEPs in XLIF. PURPOSE The purpose of this study was to assess the feasibility of using transabdominal muscle action potential (TMAPTM, NuVasive, Inc. San Diego, CA, USA) generation in lumbar fusions as an alternative to transcranial MEPs. This first analysis focuses on the feasibility and reproducibility of using this monitoring modality to activate individual lower-extremity myotomes. STUDY DESIGN/SETTING A prospective, multicenter, observational evaluation of the feasibility and utility of lower extremity nerve root compound muscle action potential monitoring in response to transabdominal stimulation was undertaken. Enrollment was completed at 15 US sites. Primary inclusion criteria included patients being at least 18 years of age who were undergoing extreme lateral interbody fusion (XLIF, NuVasive, Inc., San Diego, CA, USA) between L2-3 and L4-5, inclusive of L4-5. PATIENT SAMPLE A total of 335 patients were enrolled. Mean age of the patients was 64 years, 60% of patients were female, and patients had an average BMI of 30.6. A total of 524 levels of XLIF were treated in the 335 patients (mean, 1.6 per patient). The most common indication for surgery was spondylolisthesis (79%). OUTCOME MEASURES TMAP stimulation thresholds (mA) and myotome measurements (using a 30mV CMAP amplitude threshold) were taken prior to the initial incision at baseline, throughout the procedure at five minute intervals, and then again after closure. METHODS The technique for TMAP generation included the application of standard EMG bipolar pair recording electrodes on five myotomes per leg, one pair each on adductor magnus, vastus lateralis, vastus medialis, tibialis anterior, and biceps femoris. Stimulation was delivered transabdominally at approximately the superior end of the cauda equina. This stimulation pathway was selected to enable direct depolarization of the entire lumbosacral motor pathway as superior as possible to the surgical corridor, while remaining inferior to the spinal cord. The stimulation cathode was placed on the patient's posterior midline near the L1-2 disc space. The stimulation anode was placed on the patient's anterior midline superior to the umbilicus. For each TMAP stimulation, a series of pulses was applied with increasing stimulation intensity to identify the minimum stimulation intensity (threshold) required to elicit a compound muscle action potential (CMAP) for each individual myotome. RESULTS At baseline, all patients were stimulated transabdominally and a discrete CMAP response from at least one lower extremity myotome recorded. Out of a total of 3,350 myotomes (five bilateral myotomes per patient, per stimulation time point) stimulated at baseline, 3,255 (97.2%) yielded a CMAP in response to TMAP stimulation. At surgical closure, all patients had at least one myotome that was able to be activated with TMAP, with 94.6% of myotomes (3,169 myotomes) yielding a CMAP. Mean baseline stimulation threshold was 570mA (standard deviation (stdev), 192mA) with mean stimulation at closure of 599mA (stdev, 202mA), p CONCLUSIONS These data suggest that lower extremity nerve root monitoring can be accomplished reproducibly with TMAPs without the associated challenges or risk of traditional transcranial MEPs. While the TMAP technique using a novel stimulation application can supplement neural mapping via evoked EMG during XLIF, it also has potential utility in all lumbar surgical procedures where nerve roots are at risk. Subsequent analyses from this project will include injury detection and avoidance automation assessments.