Author

Marina Abramova, MD - LSU Health Sciences Center, New Orleans, LA

About this chapter

EMG, or electromyography, and SSEP, or somatosensory evoked potentials, are two methods of neurophysiologic monitoring. This chapter describes how they can assist in ensuring correct placement of electrodes in spinal cord stimulation. Particular attention is paid to the mention of epidural paddle electrode systems.

Index of Sections

Section 4.1: General Uses of Neurological Monitoring in Spine Surgery

Section 4.2: Laminectomy Electrode Implantation, General Methods

Section 4.3: Technique of Midline Positioning of the Spinal Cord Stimulator – Tripolar paddle

Section 4.4: New frontiers of intraoperative EMG application

Section 4.5: References

Section IV.4.1: General Uses of Neurological Monitoring in Spine Surgery

Intraoperative neurophysiological monitoring has become a routine procedure in complex spine surgery.   Somatosensory-evoked potential (SSEP) recording has been advocated to monitor the functional integrity of the nervous system during surgical manipulation [22-24, 35].  When stimulated, sensory afferents give rise to signals, carried via the dorsal columns (DC), within the spinal cord to the medial lemniscus and spinocerebellar tracts, ending in the primary somatosensory cortex [4].  SSEP monitoring does not involve the motor pathways, which in some clinical situations can lead to false-negative results and postoperative neurological deficits undetected intraoperatively  [1-3, 5, 6, 7, 12].  Dermatomal SSEP testing allows for assessment of individual nerve roots during surgery and has been shown to be more sensitive [7, 8].  However, the sensitivity and specificity of this method varies and is less well-liked than electromyographic (EMG) monitoring [8, 9].  EMG has become the standard of practice in complex spine surgery, providing surgeons with accurate feedback about individual nerve root activity during surgical manipulation of neural structures [10-14].

Section IV.4.2: Laminectomy Electrode Implantation, General Methods

For a number of conditions, particularly the “postlaminectomy syndrome” (aka “failed back”) [25, 26, 31], and Complex Regional Pain Syndrome (CRPS I and II) [31, 32], spinal cord stimulation (SCS) has been found to be one of the more effective therapies [30, 31] cost effective [27-29].  Originally, it was based, conceptually, on the gate control theory of pain introduced by Melzack and Wall in 1965 [16], which proposed that the activity of large-diameter fibers (α/β fibers) in the DC plays an inhibitory role in transmission of pain signals to the brain.  SCS subsequently received FDA approval in 1979 as a class II device for chronic pain of the trunk and limbs.  Over 50,000 SCS systems are implanted for pain disorders around the world each year.  Certain principles have become clear through the use of SCS: 1) paresthesia coverage must precisely overlap the painful areas, the cathode is the “active” contact and must deliver its energy to the DC, and 2) accurate identification of the "physiological" midline (PM) is essential for treatment of bilateral pain.  Displacement of the cathode up to 1 mm from the affected  side works well for unilateral pain [18].  Placement of stimulating electrodes in the subdural space was initially described by Shealy [17].  This resulted in a high complication rate leading to placement of the electrodes in the epidural space. Percutaneous placement of spinal cord stimulator electrodes has several advantages including: minimal invasiveness, reduced cost, and increased flexibility in lead repositioning.  Percutaneous placement of electrodes is achieved with local anesthesia and fluoroscopic image guidance to ensure the proper placement of the leads [20, 34] and is routinely used for trial electrodes. However, percutaneous placement is still associated with a comparatively high rate of migration of the electrode (up to 13.5%)[19, 20, 34, 41].  Higher power requirements with omni-directional contacts result in increased battery drain and decreased device durability when used for permanent implants.  Accordingly, many centers prefer a laminectomy approach with placement of a “paddle lead” configuration.  This approach also allows the electrode to be directly anchored to the lamina at the site of entry to the epidural space [21, 33].  Due to its invasiveness and associated patient discomfort, this approach is often performed under general anesthesia with fluoroscopic guidance to determine the anatomic midline by bony landmarks. As mentioned previously, experience has shown that outcome is critically dependent on symmetric placement of the electrode on the PM , and that the physiologic and anatomic midlines are frequently not identical, differing by up to 2 mm in up to 40% of patients [40].  In the author’s experience, it is not at all uncommon for comparative midlines to deviate significantly over as little as a single spinal level .  The importance of proper positioning of electrodes during SCS is reinforced by the computer models developed by Holsheimer et al. [39], in which they show that thickness of the dorsal CSF layer and precise midline positioning of the electrodes are the most significant factors determining the perception threshold  of stimulation-induced paresthesias.

In order to address these concerns, some centers have employed a “wake up test” to demonstrate that the paresthesia maps to the painful areas before closing.  This has received very limited acceptance because of discomfort to the patient and the difficulty obtaining trustworthy feedback from an intubated patient.  Other centers have gone to laminectomy under only local anesthesia.  Difficulty obtaining adequate intraoperative pain control has limited the acceptance of this technique.  Some centers perform the procedure under a high (midthoracic) spinal or epidural anesthetic producing a regional anesthesia, which in experienced hands, does not interfere with the distal patterns of paresthesia coverage.  However, it is prone to difficulties in execution, which can lead to dense anesthesia or significant additional intravenous sedation inhibiting patient examination.

The remainer of this chapter is devoted to an alternative technique, which allows the comfort and safety of general anesthetic, while determining the PM by objective neurophysiological testing . The identification of PM by use of evoked potentials was introduced by Claudio Feler, who obtained a patent for a device to perform the mapping [US 6,027,456].  While the device did not gain widespread usage, a few centers adopted the methodology using standard intraoperative electrophysiologic monitoring equipment.  To obtain the optimal coverage over painful areas, two major criteria must be met: the applied stimulation should be positioned longitudinally along the DC and the PM must be identified.  When general anesthesia is used, intraoperative neurophysiological monitoring with evoked potentials becomes the only way to determine the PM.  Stimulation of various portions of the dorsal spinal cord produces paresthesia in a given distribution in the conscious patient, as well as reliable patterns of sensory (SSEP) and motor unit action potentials (MUAP’s) of the EMG in the patient under anesthesia. In addition, the output data may include interpolations between specific measured points for optimal assessment of applied stimulation between evaluated lateral positions [US 6,027,456].  

These fundamental findings have been implemented in practice by the senior author (KA), who began using this approach in 1999, noting a marked improvement in outcomes over the previous flouroscopically guided technique. In this technique, MUAP’s via EMG activation are used to determine the PM by examining the symmetry of the evoked potentials with presumed midline stimulation.  In addition, it became clear that objective MUAP’s via EMG activation of specific muscles corresponded with postoperative induced paresthesia in particular regions depending on laminectomy level.  For example, EMG activation of the external oblique muscle from a T9-10 thoracic paddle consistently correlates with low back paresthesia (Figure 4.1). 

_____________________________________________________________________________________________________________________________________

These correlations are summarized in Table 1.  Additional correlations have been made for cervical SCS (Figure 2) and sacral nerve root stimulator (SNRS) (Figure 3) laminectomy placements (Tables 2 and 3).

_{Figure 4.2.  Intraoperative view of cervical paddle lead implantation. A. The electrode is behind the body of C3 and C4, entering at the C4/5 interspace.  B. With left sided stimulation, there is left side triceps activation, which will correlate to a C7 dermatomal paresthesia.}

_________________________________________________________________________________________________________________________

_{Figure 4.3.  Permanent retrograde implantation of sacral root paddle leads.  A.  Stimulation is left sided, with the cathode at the second position and the anode at the third.  B.  In this older tracing, the stimulation in the second left contact produces primarily adductor hallucis activation, solely on the left side.  This correlated with the postoperative paresthesia felt in the S3 perineal region.}

_{__________________________________________________________________________________________________________________________________________________________}

Section IV.4.3: Technique of Midline Positioning of the Spinal Cord Stimulator – Tripolar paddle

Once the 3-column paddle is placed in the dorsal epidural space, the superior midline contact is stimulated at minimal settings and the EMG trace recording associated with the dermatomal level of stimulation is monitored.  The stimulus intensity is gradually increased until MUAP’s are seen on EMG.  The lowest stimulus intensity needed to elicit a motor response is referred to as the threshold stimulus.   MUAP’s will be seen bilaterally at the threshold stimulus if the midline contact of the SCS is in line with the PM of the spinal cord.  If MUAP’s are seen unilaterally, the threshold intensity for that side is recorded and the stimulus is further increased to elicit a response on the other side.  A difference in threshold stimulus intensity between the left and right sides indicates the SCS is lateral to the PM.  However, medial/lateral repositioning of the SCS is necessary only if the difference in threshold intensity between the two sides is greater than 2 mA.  In this case the paddle may not be perfectly flat on the lateral X-ray (Figure 4.4A), thus dissection of the lateral recesses or proximal/superior lamina is further performed until the electrode is perfectly aligned with vertebrae on a lateral view (Figure 4.4B).  This assures optimal electrode column symmetry and programmability of the PM.  Of course, as a last resort, the laminotomy can be extended to a full laminectomy to allow perfect alignment of the paddle on the PM.  In this case, tissue must typically be identified to suture the electrode in place to prevent migration.  These tenants hold true for all paddle (1, 2, 3 and 5 column) array configurations.

Section IV.4.4: New frontiers of intraoperative EMG application

Section IV.4.5: References