Introduction

Cranial forms of neuromodulation for pain include deep brain stimulation (DBS), motor cortex stimulation (MCS), and non-invasive forms of electrical stimulation such as transcranial magnetic stimulation and transcranial direct-current stimulation.¹ DBS and MCS involve the surgical implantation of electrodes within the brain or in the epidural space, respectively. The DBS and MCS electrodes are then connected subcutaneously to an implantable pulse generator (IPG), most often located in the chest. As with other forms of surgical neuromodulation, careful patient selection is critical to the success of DBS and MCS for pain, and a trial of externally driven stimulation is typically performed before placement of the IPG.

Neither DBS for pain nor MCS for pain are performed with great frequency anywhere in the world. As a result, published accounts of these procedures are primarily in the form of case reports and case series. In addition, clinically significant complications in these procedures are rare, in the range of 1% to 10%. For this reason, precise estimates of rates of complications are difficult to determine for either DBS or MCS for pain, although the nature of these complications has been described well. In this chapter, the clinically significant complications of DBS and MCS are described, with emphasis on some strategies used in neurosurgical practice to avoid these adverse outcomes.

Deep Brain Stimulation for Pain

Deep brain stimulation for medically refractory pain was the first application of chronic intracranial DBS. In 1973 Hosobuchi et al² implanted a stimulating electrode into the ventroposteromedial (VPM) thalamus to treat facial pain. Chronic stimulation was attempted after the observation by Hosobuchi and others that acute stimulation before lesion placement resulted in pain improvement. Today, DBS for medically intractable pain typically targets two regions, the sensory ventroposterolateral (VPL)/VPM thalamus and the periaqueductal/periventricular (PAG/PVG) grey matter. Common indications for DBS for pain include chronic poststroke pain syndromes and chronic facial pain syndromes. Rates of reported efficacy for these procedures vary widely, ranging from 12% to 60%. The U.S. Food and Drug Administration (FDA) initially approved and then rescinded the approval of DBS for pain.

The clinically significant complications of DBS take the form of surgical complications, hardware-related complications, and stimulation-dependent complications.³ Surgical complications of DBS surgery include intracranial hemorrhage and electrode misplacement. Hardware-related complications include electrode fracture, electrode erosion, and infection. Stimulation-dependent complications are the effect of undesired modulation of neural circuits adjacent to the targets of neuromodulation.

Surgical Complications and Avoidance

Intracranial Hemorrhage

Symptomatic intracranial hemorrhage is the most feared complication of DBS surgery, with an incidence of approximately 1% of patients undergoing DBS surgery. A number of source factors may influence hemorrhage rates. Bleeding may occur because of direct trauma at the brain surface; injury to vessels in cortical sulci; or injury to deeper vascular-rich structures such as the ependymal surface, the choroid plexus, or friable target regions.

During surgery, careful planning is performed to avoid traversing cortical sulci because of the presence of vessels in the subarachnoid space. If the planned trajectory traverses the ventricles, the surgeon is obliged to determine that the electrode path does not pass through the location of large veins, such as the thalamostriate veins, which are located along the caudate head, or through well-vascularized structures such as the choroid plexus. Disruption of the ependymal surface through a trajectory that skims the surface of the ventricle over a distance may also be a source of intraventricular hemorrhage.

In the author’s center, multiple steps are taken to minimize the risk of hemorrhage. Meticulous surgical planning is performed to plan both entry points and trajectories for DBS placement, as outlined above. Verification of positioning is performed with the minimum number of microelectrode passages, as rates of hemorrhage are believed to scale with approximately 0.2% per electrode track. Systolic blood pressure is carefully monitored and routinely maintained below 150 mm Hg. For patients taking chronic antithrombolytic therapies, aspirin and Coumadin are halted 1 week before DBS lead placement and remain off until 1 week after DBS placement.

Venous Infarction and Air Embolism

Venous infarction with delayed hemorrhage is another potential complication of DBS surgery. Burr holes for DBS are placed in close proximity to the midline in the region of the coronal suture, where large cortical veins often enter the superior sagittal sinus. Disruption of these vessels or inadvertent cauterization of large veins may lead to compromised venous drainage, venous engorgement, hemorrhage, or venous infarction.

A second form of venous complication during surgery is venous air embolism. DBS surgery is most often performed in a sitting position, with the cranial opening above the level of the heart. As a result, a potential for venous air embolism exists. The rate of asymptomatic and symptomatic air embolism is not known, although coughing and transient cardiovascular changes are occasionally observed in DBS surgery.

Venous complications of DBS surgery may be minimized through planning and surgical technique. Injury to cortical veins may be avoided through the administration of contrast during preoperative imaging so that burr hole locations may be placed distant to venous confluences. If a burr hole is placed directly above a large collection of veins, a new burr hole may be placed rather than risking injury to important surface veins. In addition, meticulous attention during dural opening potentially allows entry into dural venous lakes and inadvertent injury to superficial veins. In the author’s center, venous air embolism is treated prophylactically with the administration of 500 mL of normal saline to increase venous pressure at the start of the case. In addition, bone edges, where air may enter diploic veins, are waxed carefully and quickly during opening. If signs of venous air embolism are observed during surgery, irrigation of the field and, rarely, placement of the patient into a head-lowered position can often avoid worsening of the patient’s cardiopulmonary status.

Electrode Misplacement

Inaccurate placement of the DBS electrodes into a location where there is diminished clinical efficacy or undesirable side effects is a second concern. DBS for pain involves lead placement into deep structures through a burr hole in the skull. Because direct visualization of the target is not observable, the method of stereotactic placement is critical to placement accuracy. The most common targeting modality is frame-based stereotaxy. In this approach, a frame is affixed to the patient’s skull under local anesthetic, and stereotactic magnetic resonance imaging (MRI) is performed. This MRI, together with the frame components, allows a one-to-one mapping of each location in the brain to a unique set of coordinates. Common frame systems are the Leksell Frame, the CRW Frame, the NexFrame, and STarFix systems. These systems achieve targeting accuracy by rigidly attaching the targeting mechanism to the skull. An alternative approach is to use intraoperative MRI to observe the location of the DBS lead as it is placed into the brain.

Fluoroscopic visualization allows verification with some systems. In the Leksell system, a pair of targets allows visualization of electrode location with respect to the target (Fig. 4-1). The position of the electrode can then be corrected before patient testing.

Fig. 4-1 Lateral view of thalamic deep brain stimulation (DBS). The four-contact DBS electrode located within the thalamus is secured at the skull level with a radiolucent burr hole cap. The lead terminates at a second four-contact connector to which the extension lead is attached. The extension lead then passes subcutaneously down the neck to the chest.

In addition to accurate frames, microelectrode recording can be used. Microelectrode recording allows for measuring the responsiveness of individual neurons in the brain. During placement of DBS electrodes in the VPM/VPL thalamus, recording of cellular activity while stimulating a target region of the body can confirm localization. In addition, microstimulation, applying a stimulus through the microelectrode with a patient awake, can be used to confirm that the lead trajectory is desired.

Motor Cortex Stimulation for Pain

Motor cortex stimulation for pain involves the surgical placement of stimulating electrodes in the epidural space overlying the primary motor cortex (precentral gyrus). Application of stimulation to the sensory cortex (postcentral gyrus) alone is ineffective in the treatment of pain. The paddle leads used are those designed for epidural spinal placement and may be oriented either along the precentral gyrus, parallel to the central sulcus, or bridging the precentral and postcentral gyri perpendicular to the central sulcus. MCS for the treatment of thalamic pain syndrome was reported first by Tsubokawa et al⁴ in 1991. Today the most common indications for MCS are trigeminal neuropathic pain and thalamic pain syndromes, although the therapy has been applied on a more limited basis to many other painful processes. For thalamic poststroke pain, approximately two-thirds of patients are reported to achieve adequate relief from the therapy. For trigeminal neuropathic pain, impressive response rates of up to 75% to 100% have been reported in some studies. As with DBS for pain, the FDA has not approved MCS as a therapy for pain, reflecting the difficulties of patient selection and treatment efficacy for this therapy.

MCS has appeal for its reduced invasiveness compared with other intracranial ablative and neuromodulation surgeries. The majority of studies of MCS have reported no complications of the therapy. However, occasional complications are known to occur. As with DBS, clinically significant complications may be categorized as surgical, hardware related, and stimulation related. Surgical complications of MCS involve epidural hematoma and electrode misplacement. Hardware complications include infection and fracture or dislocation of the lead. Stimulation-related complications of MCS include undesirable paresthesias and seizures.

Surgical Complications and Avoidance

Epidural Hematoma

The postoperative accumulation of blood in the epidural space can produce mass effect on the brain and compromise stimulation efficiency and efficacy. Two surgical techniques are used for the placement of MCS electrodes into the epidural space. In the first approach, a craniotomy that is 4 to 5 cm in diameter is created, with removal of the bone flap and care taken to strip the dura from the underside of the bone flap during elevation to avoid dural tearing. During the procedure, the MCS leads are secured to the dura, and the bone is replaced atop the leads and then secured into position with titanium plates. In the second (and less invasive) approach, a burr hole is placed to access the epidural space, and a dural separator is used to expand a small epidural pocket adjacent to the burr hole into which the lead may be placed.

To minimize the development of postoperative epidural hematoma, care is taken to reduce the size of the epidural potential space. In the case of craniotomy, dural tack-up sutures may be placed to reapproximate the dura to the bone flap during closure. When burr holes are placed, limiting dural stripping to the size of the implanted electrode, so that the electrode itself fills the potential space, is believed to reduce the incidence of epidural hematoma. A postoperative computed tomography (CT) scan 4 to 8 hours after the surgery can be a useful tool to evaluate the accumulation of any epidural blood.

Electrode Misplacement

As in any other form of neuromodulation, the anatomic placement of the electrode is of critical importance to the success of the therapy. MCS is no exception. What is difficult in MCS is that the optimal electrode location is not known with certainty. In practice, MCS electrodes are targeted to cover the portion of the anatomic homunculus where the patient is experiencing pain. However, the relationship of electrode placement to therapeutic efficacy remains a subject of study and discussion.

Several techniques allow identification of the sensory and motor cortex during surgery. In the author’s center, a stereotactic MRI and frameless stereotaxy are used to identify the location of the central sulcus. With a frameless imaging system, the location of the electrode with respect to the underlying brain may be directly evaluated. Ultrasound imaging of the cortical surface may also be used if an appropriately sized ultrasound probe is available. When a burr hole is used and therefore the location of the electrode cannot be directly visualized, fluoroscopy can be used to evaluate the precise location of the epidural lead (Fig. 4-2).

Fig. 4-2 Lateral view of motor cortex stimulation. Two four-contact paddle leads are placed through adjacent burr holes and located in the epidural space with the anterior two contacts overlying the primary motor cortex and the posterior two contacts overlying the primary sensory cortex. Somatotopically, the superior lead is overlying the hand area, and the inferior lead is overlying the face area in this patient with centralized trigeminal pain.

As in DBS surgery, electrophysiologic recording can also provide insight into physiologic lead location. Electrophysiologic pulsatile stimulation of the median nerve of the contralateral upper extremity evokes a negative postcentral N20 wave and a positive precentral P20 evoked potential.⁵ These sensory-evoked potentials can be used reliably to identify the location of the electrophysiologic border between sensory and motor cortical regions, allowing the lead to be placed over the motor cortex. A volumetric postoperative CT scan co-registered to a preoperative MRI can additionally allow simultaneous visualization of the brain and electrode contacts, verifying electrode locations with respect to the brain surface.

Conclusion

Deep brain stimulation and MSC are cranial neuromodulation options in the treatment of pain. Neither approach carries the approval of the FDA, and the rare complications of these procedures can be significant. However, in carefully selected patients who have exhausted other options, both DBS and MCS may be reasonable therapeutic strategies for the treatment of medically intractable pain syndromes.