Dynamically evoked, discrete-threshold electromyography in the extreme lateral interbody fusion approach

because the psoas muscle, which contains nerves of the lumbar plexus, is traversed during the extreme lateral interbody fusion (XLIF) approach, appropriate nerve monitoring is needed to avoid nerve injury during surgery and prevent approach-related neural deficit. This study was performed to assess the effectiveness of dynamically evoked electromyography (EMG) to detect and prevent neural injury during the XLIF approach.

one hundred two patients undergoing XLIF at L3-4 and/or L4-5 were enrolled in a prospective, multicenter, nonrandomized clinical study. The EMG threshold values for each of the 3 successive dilators were recorded at the surface of the psoas muscle, mid-psoas, and on the spine. At each location, the dilators were rotated 360°, taking recordings immediately posterior, superior, anterior, and inferior. For each dilator, the authors noted the rotational position (the angle in degrees) at which the lowest threshold was found. Findings of pre- and postoperative neurological examinations were also recorded.

nerves were identified within proximity of the dilators (alert-level EMG feedback) in 55.7% of all cases during the XLIF approach. Although nerves were more commonly identified in the posterior margin (63%), there was significant variability in the location of nerves identified. Despite the fact that the posterior half of the disc space was targeted in 90% of cases, no significant long-lasting neural deficits were identified in any case; 27.5% experienced new iliopsoas/hip flexion weakness and 17.6% experienced new postoperative upper medial thigh sensory loss. Transient motor deficits were identified in 3 patients (2.9%), and all had resolved by the 6-month follow-up visit.

the ability to identify and report a discrete, real-time EMG threshold during the transpsoas approach helps to avoid nerve injury and is required for the safe performance of the XLIF procedure. Additionally, nerve location is variable, thus reinforcing the need for real-time directional and proximity information.