Motor Cortex Stimulation

Published on 26/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2752 times

CHAPTER 168 Motor Cortex Stimulation

Jaimie M. Henderson, Shivanand P. Lad, Kevin Chao

Electrical stimulation of the nervous system has been known since antiquity to have analgesic effects. Intermittent application of electricity for the treatment of pain has a long and colorful history, culminating in the development of modern systems for the chronic administration of therapeutic electrical current.¹ Following the introduction of spinal cord stimulation in 1967, electricity has been used successfully to modulate pain by stimulation of peripheral nerves and central nervous system structures, including the spinal cord, brainstem, thalamus, and most recently the cerebral cortex.

In the early 1990s, motor cortex stimulation (MCS) was introduced as a treatment for central deafferentation pain. This severe pain syndrome occurs with interruption of the spinothalamic pathway, most commonly following hemorrhagic or ischemic stroke. The primary sensory cortex was initially targeted as the final relay in the spinothalamic system; however, it was soon realized that stimulation of the motor cortex was more effective in controlling pain.² Chronic stimulation of the precentral cortex for the treatment of pain was first introduced by Tsubokawa and colleagues in 1991.³^,⁴ Subsequently, there has been growing corroborative evidence derived from case reports and individual case series to suggest that epidural MCS can be an effective treatment for many patients with intractable neuropathic pain. Poststroke pain, phantom limb pain, multiple sclerosis, spinal cord injury pain, postherpetic neuralgia, and neuropathic pain of the limbs or face have all been reported to respond favorably to MCS.⁵ Most clinical studies focus on the use of MCS in poststroke pain, for which there are few other treatments. Poststroke pain responds variably to MCS, with about 50% of patients achieving adequate relief. However, some studies have documented excellent results in the treatment of trigeminal neuropathic pain, with 75% to 100% of patients achieving good or excellent pain relief.⁶^–¹⁰ MCS thus appears to hold promise for patients with intractable and otherwise difficult-to-treat pain syndromes.

Surgical Technique

Following Tsubokawa’s initial report,⁴ several studies were published using a similar technique of introducing an epidural electrode through a bur hole under local anesthetic.^3,^6,^8,¹¹^–¹⁵ In several cases, groups that started out using a bur hole technique later switched to performing electrode placement through craniotomy,^6,^8,^9,¹⁴ and most surgeons now perform a craniotomy either under local^{6–8,10,13,14,16–21} or general anesthesia.²²^–²⁹ Nearly all investigators place the electrodes epidurally, although subdural placement has been described.³⁰^,³¹ Image-guided neuronavigation is used to precisely identify the motor cortex intraoperatively,^9,^20,²⁷^–²⁹ ^,32 and proper placement is confirmed with physiologic testing.

Fiducials are placed preoperatively on the scalp, and volumetric magnetic resonance imaging (MRI) is obtained. The central sulcus, sylvian fissure, and inferior and superior frontal sulci are identified. The “hand knob,” a distinctive area of motor cortex, can usually be easily identified in oblique sections (Fig. 168-1). For facial pain, the optimal target is often identified anterior to the central sulcus adjacent to or below the inferior frontal sulcus. Some investigators have used functional MRI (fMRI) to assist in identifying the appropriate areas for stimulation (Fig. 168-2). Prophylactic antibiotic medication is given along with anticonvulsant agents. The patient is placed in a supine position with a roll beneath the shoulder and the head rotated to the ipsilateral side of the pain. The target is mapped onto the contralateral scalp, and an incision centered over the central sulcus is made.

FIGURE 168-1 Screen shot from the surgical navigation system demonstrating localization of the central sulcus and primary motor cortex. The white arrow indicates the hand knob of the motor cortex just anterior to the central sulcus.

FIGURE 168-2 Three-dimensional reconstruction of a functional magnetic resonance imaging (fMRI) study showing activation of the hand area of the motor cortex. Final electrode contact positions are indicated by the small numbered colored spheres.

After the craniotomy flap is made, a diagnostic electrode array is placed in the epidural space aligned perpendicular to the expected position of the central sulcus, covering both the regions of the precentral and postcentral gyri. Median nerve somatosensory evoked potential may be obtained to identify the N20/P20 phase reversal that occurs across the central sulcus. This provides electrophysiologic confirmation of the location of the hand region of the precentral gyrus, located cephalad and medial to the facial region. The target for electrode implantation is marked on the dural surface. Stimulation is performed in a train of three at 200-microsecond durations with currents of up to 20 mA. The painful area contralateral to the site of stimulation is observed for muscle contractions during test stimulation, and the threshold for motor response is noted. Electromyography electrodes may be placed in appropriate muscle groups to monitor for muscular contractions. Iced saline should be kept available to irrigate the brain in case a seizure is induced.

When the appropriate cortical target has been confirmed, the diagnostic grid is removed, and a 4-contact paddle electrode is positioned over the motor cortex, parallel to the central sulcus. Some investigators have described placement of the electrode perpendicular to the central sulcus, with the two central contacts of the electrode array over the target point in the motor cortex. A second electrode may be placed posteriorly over the central sulcus to allow for transsulcal stimulation and increased programming options (Fig. 168-3). Each electrode is sutured to the outer layer of the dura. If an externalized trial is to be performed, the lead wire is tunneled out through a separate stab incision, and the craniotomy bone flap is secured.

FIGURE 168-3 Postoperative skull film with two parallel leads in place. One lead is targeted to the motor cortex, and the second lead is placed immediately behind it over the central sulcus near the sensory cortex, allowing transsulcal stimulation patterns to be programmed.

The electrode is tested once the patient is fully alert. Testing may continue for several days or until the patient can consistently confirm that stimulation reduces the preoperative pain by at least 50%. If the trial is successful, an implantable pulse generator is placed in the upper chest wall under general anesthesia and attached to the leads with extension cables.

Stimulation Parameters

There is tremendous variation in reported stimulation parameters for MCS. Pain relief can occur at amplitudes from 0.5 to 10 V, rates from 5 to 130 Hz, and pulse widths from 60 to 450 microseconds.³³ Although most studies have used rates of about 40 Hz, others have found higher rates to be necessary in some cases. There is also no agreement on whether wide or narrow pulse widths provide more effective stimulation. Amplitudes have in many cases been empirically chosen, whereas other investigators base stimulation amplitude on a percentage of motor threshold.

Pain relief is most commonly achieved at amplitudes of 6 V or less, with average amplitudes of 5 V or less in most studies. Amplitudes greater than 6 V are more likely to be associated with seizures during programming, with seizures commonly induced at amplitudes approaching 9 V.³⁴

Many investigators have noted that MCS frequently produces a period of poststimulus pain relief that can range from minutes to hours. Thus, most publications report the use of a cycling mode of stimulation, with 10 minutes to 3 hours on stimulation followed by 15 minutes to 6 hours off stimulation. In one study, switching from a continuous to cycling mode (in addition to other programming changes) may have contributed to improvement in pain relief.³⁴

Indications and Results

Central Deafferentation Pain

MCS was originally developed in response to the lack of effective treatments for central deafferentation pain. In Tsubokawa’s first published report, 5 of 12 patients had complete resolution of their pain at 1 year of follow-up, with another 3 exhibiting significant decrease in pain.³ These encouraging early results led many other investigators to pursue this technique for treatment of this otherwise difficult-to-treat condition.^9,^11,^12,^17,^22,^23,^28,^29,³⁵^–³⁸

Nguyen and colleagues ⁹ reported 13 patients with central deafferentation pain from ischemic and hemorrhagic stroke as well as trauma (1 case) and abscess (1 case). Ten patients (77%) had “significant improvement,” classified as more than 40% reduction in pain. In a large series of 31 patients with poststroke pain (3 bulbar, 20 thalamic, and 8 suprathalamic), Katayama and associates¹⁷ obtained “satisfactory” pain relief in 15 (48%) with follow-up periods of more than 2 years. In 20 patients with central neuropathic pain, including 16 with poststroke pain, Mertens and colleagues achieved “excellent” pain relief in 25% and “good” pain relief in 35%. Other series have generally corroborated these results, with about 50% to 60% of patients achieving significant pain relief.

Trigeminal Neuropathic Pain

Trigeminal neuropathic facial pain is a syndrome of severe, constant facial pain related to disease of, or injury to, the trigeminal nerve or ganglion. Causes of trigeminal neuropathic pain can include injury from sinus or dental surgery, skull or facial trauma, or intentional destruction for therapeutic reasons (deafferentation), as well as intrinsic pathology of any part of the trigeminal system.³⁹ Patients who fail surgical treatment for trigeminal neuralgia may sometimes develop trigeminal neuropathic pain (also called trigeminal deafferentation pain),³⁹ for which there are few, if any, effective treatments. Few significant advances have occurred in pharmacologic therapy, and anticonvulsant and antidepressant medications remain the mainstay of treatment. Deep brain stimulation of well-defined targets in the sensory thalamus and periaqueductal or periventricular gray matter has had generally disappointing results.⁴⁰

MCS has shown some promise in the treatment of trigeminal neuropathic pain, beginning with the 1993 publication by Meyerson and colleagues.⁶ In a group of 10 patients with different types of neuropathic pain, all 5 patients with trigeminal neuropathic pain obtained between 60% and 90% pain relief at 8 to 28 months. A follow-up study by Herregodts and associates⁴¹ showed 50% to 100% reduction in visual analog scale pain scores in 4 of 5 patients with trigeminal neuropathic pain.

In a 1996 study by Ebel and coworkers,⁷ 7 patients with trigeminal neuropathic pain of various etiologies were treated with motor cortex stimulation. Six of the 7 patients underwent permanent implantation, with 5 of these 6 achieving 80% or greater pain relief. Two patients subsequently lost pain relief over the course of several months, leaving 3 of 6 patients (50%) with a satisfactory result at last follow-up.

Nguyen and colleagues have published several descriptions of their surgical technique and programming approach.^8,^9,¹⁴ In their series, all patients with neuropathic facial pain achieved 40% to 100% pain relief.^8,^9,¹⁴

More recently, Brown and Pilitsis ²⁷ reported on 10 patients who underwent a trial of MCS for facial pain of various etiologies, including trigeminal neuropathic pain, postherpetic neuralgia, and central poststroke pain. All 8 patients with a peripheral neuropathic mechanism for their pain underwent placement of a permanent system after successful trial. Eighty-eight percent of these patients obtained immediate pain relief of at least 50%, and 75% experienced sustained reduction in pain at 3 to 24 months of follow-up. A review of the literature corroborated these results, showing 29 (76%) of 38 patients with neuropathic facial pain achieved at least 50% pain relief.²⁷ All patients in the implanted cohort experienced a decrease in their medication requirements by more than 50%.

Other Indications and Overall Efficacy

MCS has been used for a number of other neuropathic pain conditions including phantom limb pain,^22,^23,⁴² spinal cord injury pain,^9,^29,^32,⁴³ peripheral nerve injury,^6,^9,^22,^28,²⁹ and others. Given that only a few patients have been reported for each, no clear conclusions can be drawn regarding efficacy of MCS for these indications.

Two recent structured literature reviews have been performed to address the question of overall efficacy of MCS. In a search of the medical literature, Fontaine and colleagues ⁴⁴ found 244 references, 14 of which were of sufficient quality to subject to further analysis. Of 210 patients in these 14 studies, 43% achieved more than 50% pain relief, 57% had a “good postoperative outcome” (pain relief greater than 40% or 50%, depending on the study), and 30% had more than 70% pain relief. In central pain, 54% of patients achieved more than 40% to 50% pain relief, whereas for trigeminal neuropathic pain, 68% achieved more than 40% to 50% pain relief.

Another review published contemporaneously analyzed 22 articles and found that the weighted responder rate (responders defined as patients who showed a global response according to each study’s definition) was 72.6%.⁴⁵ However, the selection criteria for these studies was less rigorous than the review of Fontaine and colleagues, and the definition of responder is not entirely clear. Despite these limitations, it is evident that between 30% and 60% of patients achieve more than 50% pain relief with MCS.

A single prospective study incorporating a blinded evaluation period has been performed.⁴⁶ In this study, 11 patients underwent trial MCS implantation, of which 8 went on to permanent implant. These 8 patients achieved a median reduction in pain of 63.2%. Only one patient failed to achieve more than 50% pain relief. Pain scores increased to near the preoperative baseline in all patients during blinded deactivation of the stimulator. As in other studies, the presence of complete deafferentation was a poor prognostic indicator for MCS efficacy.

Safety

MCS, like any neurosurgical procedure, can be associated with risks and complications, including bleeding, infection, and neurological deficits. Although many studies have reported no adverse events with MCS,^3,^4,^11,^12,^15,^31,^42,^46,⁴⁷ there have been some serious complications reported in the literature. Two epidural hematomas have been reported, one small and asymptomatic,⁸ the other requiring evacuation and associated with persistent dysphasia.⁶ One group reported two patients with devastating cerebral hemorrhages, with one patient dying from this complication and the other remaining in a persistent vegetative state.²⁵^,³⁰ It is possible that these complications resulted from this group’s use of subdural rather than epidural placement. Infection of the hardware requiring removal or treatment with antibiotics, or both, has been reported in a number of studies.^9,^22,²⁷^–³⁰ ^,38 Some patients have also experienced wound dehiscence that resolved with surgical revision.⁹ Breakage of the hardware can also occur.²² Two patients in one study experienced transient postoperative neurological deficits (one speech, one motor).²⁹

Programming of MCS systems, as well as long-term treatment with MCS, can also be associated with certain risks and side effects. Foremost among these is the risk for seizures, which have been frequently reported. These have been variously described as brief focal seizures during programming,²⁸^,²⁹ unspecified seizures during programming,^20,^23,^25,³⁴ prolonged focal seizure with postictal speech arrest,⁷ short-lasting generalized seizures during programming¹⁷ (occurring in most patients in one study,)⁶ and generalized seizures with activation of the stimulator.¹⁰^,³⁸ In one recent study of intensive MCS reprogramming, the average threshold for inducing seizures was 8.9 V.³⁴ There is at least one patient who developed severe epilepsy after long-term MCS.¹³ To further investigate the epileptogenic potential of chronic MCS, Bezard and associates¹³ undertook a study in three macaque monkeys who were implanted with MCS electrodes. They found that with stimulation at a rate of 40 Hz and a pulse width of 90 microseconds, no seizures occurred even at stimulus intensities up to 3 mA greater than the motor threshold. Higher frequencies and pulse widths induced muscle twitching at lower amplitudes and consequently also induced seizures at lower amplitudes.

Other reported side effects from stimulation include painful stimulation of the dura mater,^6,^9,¹⁶ stimulation-induced dysesthesiae,^8,^17,⁴⁸ dysarthria,²² and fatigue.²² There are two case reports of unusual side effects: impairment in a motor imagery task⁴⁹ and development of a painful supernumerary phantom arm in a poststroke pain patient.⁵⁰ In addition, there is a suggestion that MCS may adversely affect cognitive function, especially in older patients.⁵¹

Mechanism of Action

The exact mechanism of action of MCS has yet to be elucidated and is the subject of current study. Positron emission tomography (PET) studies demonstrate that cortical stimulation increases cerebral blood flow (CBF) in the ipsilateral thalamus, cingulate gyrus, orbitofrontal cortex, and brainstem.^37,^52,⁵³ Some studies have found that regional CBF increases occur in the ipsilateral ventrolateral thalamus, which largely carries corticothalamic connections from the motor and premotor areas.^37,^52,⁵³ Regional CBF increases in this site, however, are less than those seen in the anterior cingulate gyrus, insula, and brainstem. There appears to be a correlation with the extent of pain relief and the increase in cingulate blood flow.^37,^52,⁵³ Activation of the brainstem periaqueductal gray area has also been suggested to play a role. The results of these CBF studies suggest that the somatosensory cortex is not activated by MCS, nor does there appear to be activation of the downstream motor pathways below the stimulating electrode. However, the exact effects seen may vary depending on the surgical technique and stimulation parameters chosen.