Introduction

Neurosurgical treatments for pain have historically fallen under the headings of anatomic, ablative, or augmentative, and before the gate control theory was proposed by Melzack and Wall,¹ little attention was directed toward augmentative therapies. Spinal cord stimulation (SCS) has become one of the most useful interventional therapies to treat neuropathic and ischemic pain since its introduction as an intrathecal device by Shealy et al² in 1967.

Current estimates suggest that more than 27,000 spinal cord stimulator implants were performed in 2007 alone in the United States,³ and this number will likely increase as SCS is used earlier in treatment algorithms for neuropathic pain,⁴ supported by cost effectiveness and successful outcomes.⁵

One can appreciate that increased use of the technique also increases the incidence and prevalence of associated complications, which range from the mundane to the devastating. Recent reviews place the overall incidence of complications associated with spinal cord stimulation to be approximately between 14% and 43%.⁶ Cameron⁷ performed an extensive review of the literature extending over 3679 patients, concluding an overall mean complication rate with SCS of 36%. Moreover, Kumar et al⁸ reported that complications requiring revisions were approximately 25% to 33% and approximately 85% of reoperation patients were satisfied with the results.⁹ Furthermore, Rosenow et al¹⁰ retrospectively determined that in 289 patients who underwent spinal cord stimulator implants, 46% required revision, and of those, almost half required more than one.

It should be noted, however, that the innate nature and limitations of current SCS technology requires vigilance and, as a consequence of changes in impedance and battery life, either reprogramming or surgical revision. It is far from a “just set it and forget it” device, and although adjustments need to be made, the incidence of complications dramatically decreases after the first 12 months.

As with any treatment plan, complication avoidance begins with patient selection. Patients with local infection near the injection site, coagulopathy, allergy to injectate, or comorbidities or conditions that prevent fluoroscopic needle guidance or consent should be avoided. Patient selection regarding psychometric testing deserves special mention. It is well established that concurrent psychiatric illness reduces interventional treatment success rates ¹¹ and that approximately 20% to 45% of pain patients have accompanying psychopathology.¹² Therefore it is essential that appropriate measures be taken to diagnose, treat, or exclude unsuitable candidates. Instruments described to aid in identifying the presence of clinically significant psychopathology include the Symptom Checklist 90 (SCL-90-R) and the Minnesota Multiphasic Personality Inventory (MMPI-2). Poor treatment outcome was identified in patients with presurgical somatization, depression, anxiety, and poor coping.¹³

It is also crucial to appropriately use the spinal cord stimulator for the appropriate indications (neuropathic pain states and ischemic pain), as approved by the Food and Drug Administration. Disease states that yield themselves to successful treatment with SCS include postherpetic neuralgia, intercostal neuralgia, postlaminectomy pain, complex regional pain syndrome, phantom limb pain, angina, painful peripheral neuropathy, focal peripheral neuropathies, ischemic limb pain, and radiculitis and radiculopathy.¹⁴

A clear understanding of spinal anatomy, utilization of image guidance, and proper surgical technique are vital to ensure both quality treatment and reduced patient morbidity and mortality.¹⁵ Furthermore, it should be understood that appropriate training within Accreditation Council for Graduate Medical Education accredited programs and mentorship is pivotal to ensure treatment success and limit iatrogenic morbidity; interventional hobbyists and weekend crusaders only serve to undermine the accessibility of this valuable therapy to patients.

Spinal cord stimulator complications can be broadly divided into those that are technical or those that are biologic or surgical in nature. The consequences of these complications can range from the mundane to the devastating. Most complications of SCS do not impact patient mortality,⁷ unlike that for intrathecal drug delivery, as recently reported by Coffey et al.¹⁶ Although the range of various complications varies from 0% to 81% across studies,¹⁷ the frequencies of the complications in the tables in this chapter are presented as the means. Refer to Table 1-1 for reported rates of complications, revisions, and explants associated with SCS.

Table 1-1 Reported Complications of Spinal Cord Stimulation

Complication	Frequency
Overall	14%–43%^6,¹⁷
Complication requiring revision	25%–33%,⁸ 23%¹⁷
Complication requiring explant	11%¹⁷

Selected Complications

Technical Complications

Innately, the restrictions current technology places on the spinal cord stimulator device plays a role in the associated complications. Advances in system technology have largely reduced many early complications related to device failure, including loss of stimulation. Van Buyten et al ¹⁸ described such a relationship, commenting in a 5-year prospective study of 84 patients with 85 spinal cord stimulator systems, with overall 54% of patients reporting greater than 50% pain relief at 5 years. Satisfaction and employment of the patient programmer were at least 95% after stimulator therapy began (Table 1-2).

Table 1-2 Technical Complications Reported for Spinal Cord Stimulation

Complication	Frequency
Lead migration (loss of stimulation) requiring revision	11%,²¹ 13.2%²²
Lead integrity violation requiring revision	6%,²¹ 9%²²
Loose connection	0.4%⁷
Hardware malfunction, equipment failure	2.9%,⁷ 6.5%¹⁷
Battery failure	1.6%⁷
Central nervous system electrical injury from aberrant stimulator activation	Case report²⁵
Misplacement of epidural lead	Case report²⁴

Furthermore, stimulation, regardless of the choice of the numerous mechanisms proposed,¹⁴ depends on the integrity and characteristics of the “circuit created.” Spinal cord stimulator contact to target structure length varies with patient position, as described by Abejon and Feler,¹⁹ and predictively, so too do the circuit created and the resultant therapeutic stimulation. Recumbence reduced impedance and energy requirement (measured in thoracic and cervical implanted leads), and therefore stimulation was perceived to be “stronger,” although the difference was not statistically significant. Clinical implications longitudinally require further study.

Lead Migration

The most common complication of SCS is lead migration ^7,^14,^17,²⁰ and loss of therapeutic stimulation. Cameron et al⁷ reported that 27.2% of spinal cord stimulator cases were technical in nature, and of those, 87% were related to the lead. Estimates place migration incidence to be between 11% and 13%, with a greater propensity for migration in the cervical spine,⁸ likely related to greater mobility, compared with the lumbar spine.

Understandably, migration occurs when directional forces and tensile load on the electrode exceed the stabilizing forces of the anchor. These forces depend on lead-spanning mobile areas, trajectory into the epidural space, anchor orientation, the type of tissue the anchor is sutured to, suture technique, and the site of battery placement. It has been suggested that percutaneous electrodes are more prone to migration than their laminectomy counterparts and that battery location in the abdomen may be associated with less lead migration than in the gluteal region, however work by Henderson and Kumar suggest otherwise. Those studies have suggested that the closer the implanted pulse generator (IPG) is to the lead insertion site, the lower the risk of lead movement. IPG placement choices include the buttocks, midflank, abdominal wall, or infraclavicular regions.⁷ Rosenow et al,¹⁰ however, in a study of 289 patients, reported that migration was slightly higher in the surgical laminectomy lead.

Mitigating migration involves reducing iatrogenic cofounders. These include the angle of spinal cord lead entry into the epidural space, stress loops, battery site choice, and anchor or suturing technique. Bedder and Bedder ¹⁵ detail basic surgical technique and nomenclature that will not be further discussed here. Kumar et al⁸ suggest that proper anchoring and placement of the tip of the anchor into the supraspinal ligament might reduce anchor failure. Strain relief loops of approximately 1 inch in diameter, both near the anchor and at the generator site, may reduce tensile load (Fig. 1-1).

Fig. 1-1 Proper anchoring of a spinal cord stimulator lead. A, The nose of the anchor is invested within the fascia, minimizing lead migration and lead kinking. B, The nose of the anchor is superficial to the fascia, creating a potential nidus for migration and kinking.

(Modified from Kumar K, Buchser E, Linderoth B, et al: Avoiding complications from spinal cord stimulation: practical management recommendations an international panel of experts, Neuromodulation 10:24-33, 2007.)

Sometimes extensions are needed to allow placement of the battery in a desirable location from the paraspinal access. Extensions are also used when the implanted percutaneous trial method is used; when a trial is deemed successful, the extension portion can be removed distally from the externalized extension to the IPG location and a new extension can be tunneled to the subcutaneous IPG location, while maintaining the neuroaxial segment.

Lead Fracture

Lead fracture occurs approximately 5% to 9%,²¹^–²³ with contradictory reports regarding fracture rates with percutaneous versus laminectomy leads. Kumar et al²³ described percutaneous lead fracture more often than laminectomy leads; in contradiction, Rosenow et al¹⁰ reported more lead fracture with laminectomy. Clearly, lead fracture depends on lead tensile strength, the friction coefficient of the internal wires composing the lead, and a vector of force that challenges the lead integrity. Iatrogenic ways to lessen lead fracture are similar to strategies used to reduce lead migration and include appropriate anchoring, using stress loops, placement of the IPD in the abdomen,⁸ and avoidance of tunneling around mobile structures (i.e., joints).

Hardware Malfunction and Battery Failure

Exclusive to lead migration of fracture, other hardware complications occur with a frequency of 2.9% and include battery failure (exclusive to predicted nonrechargeable battery depletion) with a frequency 1.6% per Cameron et al.⁷ Turner et al¹⁷ described equipment failure rates of 6.5%.

Iatrogenic measures mitigating equipment failure again focus on good surgical skill because of diligent testing of connections and a keen understanding of the limitations of the technology. Each stimulator company’s IPG has a different recommendation regarding depth from skin, and every attempt should be made to be compliant. Readers are directed to each company’s literature for details.

Misplacement of Epidural Lead

Therapeutic stimulation requires epidural placement of the spinal cord stimulator lead. Intrathecal and subdural placement have been reported.²⁴ Troubleshooting of lead placement centers on fluoroscopic guidance and impedance testing. Anterior epidural lead placement is suggested by muscle contraction of adjacent or distal spinal nerves and painful stimulation within the appropriate expected stimulation parameters. Intrathecal lead placement is suggested by very low-impedance testing, and the hallmark of subdural stimulation is segmental, low-amplitude stimulation that is often uncomfortable to the patient. A case report of intramedullary lead placement is described in the next section.

Aberrant Stimulation

Aberrant activation of stimulation rarely occurs, although an infamous case report is often sited. Eisenberg and Waisbrod ²⁵ described a patient who had a cervical spinal cord stimulator that was activated by an antitheft device. Six months after implantation, upon entering a store, the patient felt an electrical shock in the back of his skull and fell unconscious. He awoke in the emergency department with confusion, ataxia, upper extremity intention tremor, weakness, and dysarthric speech. Initial computed tomography (CT) scan and electroencephalogram were normal, although 6 months later, with continued symptoms, follow-up CT demonstrated a left basal ganglia infarct. The authors postulate that the anti-theft device triggered a sudden bolus of current, causing damage that contributed to the injury.

Kumar et al ⁸ reported that whereas device or technical complications typically occur within 2 years of implant, biologic complications of SCS typically occur within 3 months.

Only gold members can continue reading. Log In or Register to continue