S118. A methodological reappraisal of triggered electromyography for placement of pedicular screw

Abstract

Triggered electromyography is widely used in predicting pedicle wall break after placement of pedicle screw instrumentation. However, common experience confirms that the technique fails to detect medial pedicle breaches more often than expected, the occurrence of false negative results being the most unpleasant bias. This probably results from current dispersion via the uninsulated screw and current shunt between the anode and the return electrode, causing an unpredictable discrepancy between the initial current strength and that, always lower, actually reaching the tested root. The purpose of our technique is to correct this misleading mismatch through a strictly focal root stimulation within the pedicle hole, preceding screw insertion. In 28 patients, receiving 138 lumbosacral pedicle screws during instrumented fusion, we performed a focal root stimulation by inserting a monopolar electrode fully insulated except at the tip into the pedicle hole at a depth, customized to the individual patient, ranging from 12 to 16 mm, to make the stimulating tip directly face the tested nerve root. Moreover, to prevent current shunt between the active and return electrode bypassing the tested nerve root, the return electrode was a large (25 cm2) surface electrode placed over the middle abdomen between umbilicus and xiphoid process. In all cases threshold to screw stimulation was also measured. All subjects received a post-op CT scan. Threshold values after hole stimulation in all cases in which intra-op visual inspection (when performed), post-op clinical evaluation and CT scan ruled out a pedicle medial wall breakthrough, ranged from 2.7 to 13 mA without significant difference between L4, L5 and S1. Corresponding values for screw stimulation ranged between 9.5 and 44 mA. In 5 of 7 cases of this group with threshold values at the lower range (2.7–3.3 mA), post-op CT scan, showed an apparently intact pedicle wall but a reduced distance (1 and 1.5 mm) from the screw edge to the surface of the pedicle. The four cases in whom CT scan demonstrated a pedicle break had a threshold of 2, 1.5, 2.5 and 1.8 mA, respectively. In the first 2 cases, assuming a false positive result based on an apparently normal intra-op fluoroscopy, screw was inserted, but visual inspection after laminectomy revealed a conflict between root and screw, advising a prompt screw withdrawal and redirection. In the last 2 cases screw insertion was deferred and pedicle hole was redirected. In only 2 of them, threshold to screw stimulation was abnormal (8.4 and 4.8 mA). No false negative results were observed with hole stimulation. Focal hole stimulation provides the absolute threshold values for root activation as proved by the fact that after a pedicle wall break, threshold values coincided with those commonly observed after direct root stimulation. Early detection of a misdirected pedicle route may prevent a further damage caused by insertion of the more invasive screw.