Neuromonitoring for spinal cord stimulation placement under general anesthesia

37.1

Introduction

Spinal cord stimulation (SCS) is a common therapy used to treat medically refractory neuropathic back and other limb pain syndromes . Historically, placement of SCS leads required direct interaction with the patient to ascertain where the patient felt paresthesias during stimulation. Obtaining this feedback necessitated the use of a sedative anesthetic technique and waking the patient during periods of the surgery. The patient needed to be alert enough to respond to sensations generated from stimulation without confusion or inaccuracy. Given the potential for a wide variety of responses pain patients may have during these changing levels of sedation, variable responses to anesthetics, the range of pain distributions, and often inadequate ability of the patient to differentiate their pain syndrome from the surgical “situation” or the feeling of the stimulation at the time of testing, and positional changes from cord movement and cord location in only the prone position of the surgery, these responses have been demonstrated to be unreliable or misleading . Newer multicontact paddle lead designs have allowed for a broader ability to capture pain relief even when the lead has not been “ideally” positioned during surgery or moves slightly afterward. However, these new designs still require leads to be placed in an appropriate mediolateral position relative to the specific morphology of the dorsal column fibers and entering dorsal nerve roots, at least for paresthesia-reliant systems. The center of the cord may be more than 2 mm from the canal center in 40% of patients . Fluoroscopy, on its own, has errors associated with parallax and visual alignment, which can make midline assessment difficult or misleading. In addition, the cord itself may also be rotated slightly, making one side of the dorsal column closer to the electrode, even if it is located perfectly in the midline. A prior study using a general anesthetic technique found that visually acceptable lead placement alone resulted in 1 in 6 suboptimal lead locations .

It is critical that electrodes be placed at the appropriate cranial–caudal spine level to maximize the desired pain coverage. The cranial–caudal position is easier to locate, however, given the use of trial lead information, the known segmental dermatomal distributions, and the length of the leads, which can cover 2–3 vertebral levels. As a rule of the thumb, lower back coverage is best obtained at T8, buttock and leg coverage T9–10, and foot coverage below T10. Mediolateral electrode optimization, nonetheless, remains the most important for maintaining coverage once the lead is covered in an unknown amount of fibrosis, changing the surrounding electrical characteristics in the tissue and often requiring reprogramming .

In addition, sedative techniques alone in prone patients are not without notable risks for complications. Prone, sedated patients are more at risk of losing their airway if oversedated, as access to placing an airway is much more compromised. More sedation is typically needed for tunneling and implantable pulse generator pocket creation, even if the patient is only sedated otherwise for the initial incision and dissection to the epidural space. The balance between not enough comfort for the patient and oversedation is not always straightforward, and it is not unheard of that the surgical wound needs to be emergently packed and the patient immediately turned supine to be intubated. Data from analyses of the closed claims anesthesia cases suggest that this risk could be as high as 6%, resulting in severe respiratory depression with death or brain damage .

The use of neurophysiologic mapping techniques allows these procedures to be performed under a general anesthetic technique, virtually eliminating all of the above concerns. At present there are two primary mapping techniques that are used for placement of SCS leads. The first technique, “compound muscle action potential (AP) (CMAP) activation,” is based on the antidromic activation of the alpha-motor neuron (MN) ( light green arrows in Fig. 37.1 ) through stimulation of the large Ia fibers of the dorsal column ( Fig. 37.1 ) . This stimulation, through the electrode itself, in turn, antidromically depolarizes the MN causing a CMAP in the muscle of that motor unit ( Fig. 37.2 ). It is important to note that the stimulation needed to generate the CMAP response is typical of a slightly higher amplitude than normally used postsurgery for pain therapy. This increase in stimulation amplitude is needed to overcome the anesthetic effects at the MN synapse and the fact that these Ia fibers are only a few of over 4000–12,000 afferent synapses to the MN , requiring the activation of multiple fibers to depolarize the MN. Modeling and neurophysiologic collision studies have demonstrated that the stimulation is in the dorsal columns .

The second is the “collision” technique, based on a collision of opposing APs in the dorsal column of the spinal cord. In this technique, standard somatosensory-evoked potential (SEP) cortical amplitudes are used as an indicator of laterality . The basis of this technique relies on the fact that SEPs are generated via time locked stimulus averaging, where any signal that is not time locked will act as not only noise but also cause collisions within the spinal cord, which will “knock” out the APs generated by the SEP stimulation at the median/ulnar or posterior tibial nerves. These two facts cause a reduction in the cortical SEP amplitude on the side that is being affected by the nontime locked SCS stimulus. While both techniques are utilized, a recent study comparing them suggested that the CMAP technique was more reliable and robust .

37.2

Methods