Spinal Cord Stimulation for Chronic Pain Management Implantation Techniques

Published on 23/05/2015 by admin

Filed under Physical Medicine and Rehabilitation

Last modified 22/04/2025

Print this page

rate 1 star rate 2 star rate 3 star rate 4 star rate 5 star
Your rating: none, Average: 1 (1 votes)

This article have been viewed 2754 times

CHAPTER 30 Spinal Cord Stimulation for Chronic Pain Management Implantation Techniques

INTRODUCTION

The fact that electricity might beneficially affect painful conditions has been known since antiquity.14

Spinal cord stimulation (SCS) was introduced by Shealey in 1967.5 Initially, the electrodes were placed over the dorsal columns in the subarachnoid space through a laminectomy. Subsequently, the electrodes were implanted, always through a laminectomy, between the two layers of the dura or epidurally. Some authors demonstrated efficacy of the procedure even with electrodes implanted ventrally to the spinal cord.68 This did not prove to be a practical way of conducting spinal cord stimulation and was subsequently abandoned. In 1975, Dooley described percutaneous implantation of electrodes in the dorsal epidural space. Manufacturers involved in the early stages of SCS included Medtronic, Avery, Cordis, Clinical Technology Corporation. Initially, the stimulating systems were only radiofrequency (RF)-driven passive receivers. In the mi-1970s, Cordis introduced the first pulse generator powered by a lithium battery. This was then followed by the Itrel pulse generator manufactured by Medtronic. In the first stages, stimulation was delivered through a unipolar electrode. Subsequently, bipolar arrays were made available. Many different types of percutaneous and plate-type arrays were developed. In all of them, however, the contact combinations were hardwired, and could no be reprogrammed after the pulse generator was implanted. A very important advance stemmed by the collaboration of Joseph Waltz and the Neuromed company in the early 1980s; they produced the first percutaneous quadripolar electrode with contact combinations that could be reprogrammed noninvasively through the external transmitter.9,10

In the late 1970s, there was a surge of enthusiasm for spinal cord stimulation among neurosurgeons in Europe and in the US. Thousand of patients with almost any type of painful condition were subjected to SCS. Poor patient selection, technical problems with the implanted equipment, and implantation by surgeons with minimal experience and commitment, resulted in a large number of patients with poor results; in the early 1980s the procedure fell into disrepute and was viewed with skepticism. Only a few dedicated neurosurgeons continued to apply this procedure for pain management. Gradually, in the past 10 years, the procedure has regained acceptance in the management of chronic nonmalignant pain; its role is currently firmly established in the armamentarium of the pain specialist as proven by several reports in the literature. The procedure has been acquired by other specialties, first among them anesthesiologists specializing in pain management. Other specialties have also demonstrated interest in the procedure, although to a lesser extent, such as rehabilitation medicine and orthopedic surgery. Interestingly, a large percentage of the SCS implants currently performed in Europe (particularly in Italy and Spain) are performed by vascular surgeons for the management of peripheral vascular disease.

Because of the emerging complexity of the structures being involved by the stimulation, the term ‘dorsal column stimulation,’ which was originally applied to this procedure, has been in general replaced by the term ‘spinal cord stimulation.’

Spinal cord stimulation, even though not considered an extremely technically demanding surgical procedure, commands extreme care in the details of the planning and of the execution. One has to be extremely fastidious about the correct positioning of the electrode(s), both in the longitudinal and transverse direction in the spine, about the position of the pulse generator, about the location of the subcutaneous wires and about the hook-up of the whole system. If only one of the various factors is not optimal, the effectiveness of the whole procedure might be negated.

This chapter will discuss the various available hardware solutions and some of the technical details of surgical implantation.

CURRENTLY AVAILABLE EQUIPMENT

Currently available equipment for SCS consists of electrodes, pulse generators, radio receivers, and transmitters.

Electrodes

There are two main types of electrodes, the catheter-type (otherwise commonly referred to as ‘percutaneous’ leads) and the plate-type (otherwise commonly referred to as ‘laminotomy’ or ‘surgical’ leads) (Fig. 30.1). Percutaneous electrodes are commonly used both for trial stimulation or for permanent implantation. The most commonly used electrodes are either quadri- or octopolar. The general trend is to utilize one or two quadripolar electrodes for limb pain, and one or two octopolar electrodes for axial pain. A percutaneous electrode recently introduced by Advanced Bionics (Sylmar, CA) has 16 electrical contacts. The electrodes might be connected directly to the pulse generator/receiver, or they can be connected to an intermediate subcutaneous extension which, in turn, interfaces with the pulse generator/receiver.

Plate-type electrodes require surgical implantation under direct vision (Fig. 30.2). The amount of actual bony removal varies and is often limited to a small portion of the lamina and spinous process. The simplest quadripolar plate electrode is the Medtronic Resume and Resume-TL, and the Advanced Neuromedulation Systems Lamitrode 4 (Medtronic Inc., Minneapolis, MN; Advanced Neuromedulation Systems, Plano, TX), with all four contacts arranged linearly in one paddle. The Medtronic Specify and the ANS Lamitrode 44 have eight contacts arranged in two parallel columns. Another electrode (ANS Peritrode) consists of two smaller paddles, each with two contacts; this configuration allows the surgeon to place the paddles in two different locations or with two different orientations and therefore offers a greater degree of flexibility. Plate electrodes with one or two columns of eight contacts are also available (ANS Lamitrode 88 and Lamitrode 8).

With modern technology, both types of electrodes are safe and effective ways of delivering electrical stimulation to the spinal cord. The percutaneous technique is appealing because it allows one to insert the electrode without muscle dissection and bony removal (Fig. 30.3). This is a substantial advantage when one wants to perform a trial stimulation to assess candidacy for a permanent implant. Percutaneously placed electrodes can also be advanced over several segments in the epidural space, thus allowing testing of several spinal cord levels. By inserting multiple parallel electrodes, various configuration matrices can be constructed that allow creating extremely focused electrical fields. Placement of percutaneous electrodes must be performed under fluoroscopic guidance (Fig. 30.4). This requires wearing heavy shielded garments and potentially exposes the implanting physician to non-negligible levels of radiation. The plate electrodes require open surgical intervention. Bony removal can be very limited. In the thoracic area the lower two-thirds of the spinous process and a small portion of the lamina usually have to be removed. In the cervical area, bony removal is often not necessary, and this is particularly true when placing electrodes at the C1–2 level. Most ‘laminotomy’ implants can be done through a small (1–1.5) skin incision. By advancing the electrode in a cephalad or caudal direction, one can explore at least three spinal levels in the thoracic and 4–5 in the cervical spine. (Fig. 30.4) Multiple arrays or different electrode configurations can be also constructed by utilizing more than one plate electrode. In the author’s experience, the main advantage of plate electrodes resides in their greater inherent stability in the dorsal epidural space and lesser propensity to migrate. Plate electrodes are the only option in case of previous spine surgery at the planned implant levels. The pattern of stimulation–induced paresthesiae provided by plate electrodes might be superior to the ones produced by the percutaneous electrodes. In a randomized, prospective study, North et al. proved that the performance of plate electrodes significantly exceeded that of percutaneous electrodes.10 Concordance of stimulation paresthesiae with pain was statistically better for plate electrodes. Plate electrodes are electrically more efficient. This is due to the fact that all the current is directed toward the dura instead of being dispersed circumferentially, as in the percutaneous electrodes. Plate electrodes therefore have a lower current requirement. Another advantage of plate electrodes with two columns of contacts lies in the fact that, unlike two parallel percutaneous electrodes, the relation among the electrical contacts is fixed and completely predictable. Some situations clearly command one of the two methods (i.e. a percutaneous system in the case of an outpatient percutaneous trial, or a plate electrode in the case of prior spine surgery). In most other situations, the choice is usually dictated by individual preferences and patterns of practice. A skilled implanter can usually achieve a similar stimulation matrix with either a plate or a percutaneous electrode. The differences between plate and percutaneous electrodes is likely to become more blurred with the development of miniaturized plate electrodes that can be introduced epidurally through a percutaneous device

Pulse generators/receivers

See Figure 30.5

Stimulation consists of rectangular pulses delivered to the epidural space through the electrodes. Two types of systems are currently available. The totally implantable pulse generators contain a lithium battery in the pulse generator. They are activated and controlled by outside transcutaneous telemetry and, once activated, do not require any patient input to function. They can also be turned on and off through a small magnet. Lifespan of the battery greatly varies with usage and with the utilized parameters (voltage, rate, pulse width, etc.). Most patients can expect, under average use, the battery to last 2.5–4.5 years. Available lithium-powered pulse generators allow stimulation to be given in increments of 0.1 V. and with rates up to 130 Hz. The Precision pulse generator, manufactured by Advanced Bionics, contains a rechargeable lithium battery. The battery is recharged by wearing an outside recharger while the stimulator is in use. The interval between charges varies with usage time and power requirements and can be expected to be in the order of at least several days under normal utilization. Lithium-powered pulse generators make up the majority of implanted spinal cord stimulator.

Radiofrequency-driven systems, instead, consist of a passive receiver, implanted under the skin, and the transmitter which is worn outside of the body. An antenna, which is applied to the skin in correspondence of the receiver and connected to the transmitter, transmits the stimulation signals transcutaneously. In order for the system to function, the transmitter has to contain charged alkaline batteries and the antenna must make adequate contact with the receiver. This requires the patient to wear the outside system in order to receive the stimulation. RF-driven systems can deliver stimulation with a rate up to 1400 Hz, and can be customized to deliver more power than the corresponding lithium-powered systems.

Both systems have advantages and disadvantages. The main disadvantage of the RF systems is the inconvenience of having to wear the antenna and the radio receiver. The problem might go beyond pure inconvenience in individuals who have handicapped motor function in the upper extremities and cannot properly go through all the steps required to make the external unit function properly. Other patients, particularly individuals affected by a complex regional pain syndrome type 1, might not tolerate the antenna taped to the skin. The equipment cannot be worn while swimming or showering, and severe perspiration might make proper contact of the antenna problematic. Besides these considerations, the patient has to worry about changing batteries on a regular basis and making sure that the proper coupling exists between the antenna and the receiver at all times. These inconveniences are obviated by a lithium-powered system that runs automatically without any patient intervention. The RF system is usually reserved for patients who have greater power requirements and who would have to undergo replacement of the lithium-powered pulse generator with an unacceptably high frequency. The advent of the rechargeable lithium battery might redefine this requirement and the indications for an RF system.

The distribution of the electrical fields within the intraspinal structures is affected by the position of the electrode array as well as the polarity of the individual contacts. In order to generate an electrical field, one must have at least one negative contact activated (cathode) and one positive contact activated (anode). With the ANS and Medtronic systems, each contact can be either on or off. The Advanced Bionics System, instead, allows each electrical contact to be activated in fractional increments, thus allowing an almost seamless change in the distribution of the electrical field.

WHAT STRUCTURES ARE BEING STIMULATED

The spinal canal contains several nervous and non-nervous structures that, when stimulated electrically, give rise to a variety of responses. The electrical properties of the intraspinal contents can be characterized as the ones of an nonhomogeneous conductor (Fig. 30.6) Knowledge of the different type of responses and their correlation with the underlying anatomical substrate is extremely important in implementing strategies for spinal cord stimulation.11,12

The width of the cerebrospinal (CSF) space is the most important factor in determining the stimulation parameters, particularly the perception and discomfort thresholds (Fig. 30.7). Dorsal root fibers in general have a lower stimulation threshold than dorsal column fibers and this is particularly evident with increasing thickness of the CSF layer. This is due in large part to the fact that dorsal root fibers have a very high conductivity at their entry into the spinal cord.

Stimulation of the large myelinated afferent fibers at the intraspinal level can occur in four different areas: the dorsal root, the dorsal root entry zone, the dorsal horn, and the dorsal columns. Electrical activation of these structures elicits tingling paresthesiae that are always ipsilateral to the stimulating electrode. If the stimulation voltage is increased, discomfort and pain occur. Clinically, it is extremely important to differentiate activation of the segmentary large myelinated afferents (dorsal root/entry zone/dorsal horn) versus activation of the ascending long tracts in the dorsal columns. Activation of the segmentary afferents causes paresthesiae located in the radicular dermatome at the level of the electrode. For electrodes in the thoracic area, this usually means paresthesiae along the anterior chest wall. With electrodes placed at T12 or L1 the usual pattern is paresthesiae along the anterior aspect of the thigh or in the inguinal area. In the cervical spine, paresthesiae will be elicited in various segments of the upper extremity.

The stimulation threshold for the segmentary system is lower than the one for the dorsal columns, and usually ranges 0.1–0.5 volts Differentiating between stimulation of the dorsal root, dorsal root entry zone, or dorsal horn can be exceedingly difficult. One can assume dorsal root stimulation can be expected if the electrode is placed laterally in the spinal canal. On the other hand, stimulation of the dorsal root entry zone and/or dorsal horn is more likely if the electrode is placed near midline, and segmentary paresthesiae are rapidly followed by activation of the dorsal columns with a small voltage increment. Stimulation of the longitudinal fibers of the dorsal columns is characterized by ipsilateral paresthesiae occurring in areas of the body caudal to the level of the electrode

The exact distribution of the paresthesiae varies with the level of the electrode and with some degree of interindividual variability. In some individuals, activation of the dorsal columns gives rise to a smooth tingling sensation which is uniformly distributed in all the dermatomes caudal to the implanted electrode. This is usually the hallmark of upcoming good therapeutic effects. Other individuals, fortunately a minority, perceive the stimulation with a patchy distribution, affecting separate body segments that are not connected by paresthesiae. Individuals that exhibit this pattern of distribution of the paresthesiae, unfortunately, almost never experience any meaningful pain relief.

Activation of the dorsal columns usually occurs at a threshold that is at least 0.5–1.0 volt higher than the segmentary pathway. A thicker dorsal CSF space usually favors more selective activation of the segmentary sensory system as opposed to the dorsal column fibers. It is common to observe that, initially, systems are simultaneously activated, but after a few weeks the stimulation pattern is confined to a segmentary band. For this reason electrode placement in the upper thoracic spine (where the CSF space is the widest) seldom results in satisfactory long-term stimulation of the dorsal lemniscal pathway.

When stimulating the intraspinal structures epidurally, most commonly one observes a mixture of dorsal column, dorsal root entry/zone, dorsal root stimulation. This is true particularly with electrodes placed in the low thoracic–upper lumbar area (T11–12, L1) where the spinal cord tapers into the conus medullaris and the cauda equina nerve roots are a prominent component of the intraspinal structures.

Stimulation of the motor structures results in muscle contractions. Activation of the segmentary motor system (ventral root/motor neurons) results in muscular contractions in the somatic distribution of the stimulated segment. Activation of the descending corticospinal pathways, instead, causes muscle contractions in segments caudal to the level of the electrode. With laterally placed electrodes, stimulation of the segmentary motor system can occur at a threshold equal to or lesser than the one for large myelinated afferents. This invariably results in unpleasant contractions that can completely mitigate the benefits of the stimulation. In the vast majority of patients, stimulation over the dorsal columns does not result in activation of the motor system unless the stimulation is increased to a much higher voltage. The author, however, has occasionally seen patients in whom, even with a perfectly placed midline electrode, activation of the motor system occurred simultaneously with the sensory system. In these instances, selective stimulation of the dorsal column cannot be successfully performed.

OPERATIVE TECHNIQUE

Intraoperative anesthetic management

Electrode implantation can be performed either under monitored anesthesia (local anesthetic and intravenous sedation) or under general anesthesia. Unlike SCS for movement disorders or for peripheral vascular disease, the exact distribution of the stimulation-induced paresthesiae is crucial in pain management. Consequently, it is crucial to test an awake and cooperative patient if the best results are to be achieved. When the procedure is performed under general anesthesia, one can rely only on the radiological position and on evoked motor or sensory responses. However, when the procedure is performed under general anesthesia, one cannot obtain information about the specific details of the distribution of the paresthesiae or as to whether there is concomitant motor stimulation at the therapeutic threshold. Percutaneous electrode placement is always performed with local anesthesia and intravenous sedation. Plate electrode implantation can be performed either under monitored or general anesthesia. Unless there are medical contraindications, it is the author’s preference to perform the implantation of the stimulator under general anesthesia and then to wake up the patient intraoperatively to perform the sensory testing of the electrode positioning. Percutaneous and plate electrode implantation under general anesthesia is required at the C1–2 level.

Anesthetic management during implantation performed under monitored anesthesia is of crucial importance to the success of the procedure. Working with an uncooperative patient or one who is too sedated to be able to answer questions during the testing might result in complete failure of the procedure or, even worse, in accidental neural injury. The procedure should be accomplished with minimal discomfort for the patient. The patient should be fully awake and cooperative during the testing phase. With keen attention to detail and good cooperation with the attending anesthesiologist, these goals can be satisfactorily reached in the majority of the patients.

Generous anesthetic infiltration with a long-acting agent minimizes the requirements for intravenous sedation. Whether the procedure is percutaneous or through a laminotomy, the periosteum must be generously infiltrated. The dura is often painful when first touched by an instrument or the needle. The pain is usually sharp and localized to the stimulated site. Dural pain, however, usually habituates quickly, and only minimal/moderate discomfort is usually perceived upon further dural manipulations or needle insertions. Some authors have advocated injecting local anesthetic (1% lidocaine or 0.25% bupivacaine) in the epidural space. This results in analgesia to painful pinprick and reduces local anesthetic requirements without preventing the patient from warning that the needle may be piercing the dura. This technique has been shown not to affect the electrical stimulation of the large myelinated fibers in the spinal cord and therefore does not jeopardize the reliability of the intraoperative testing.

Excessive use of intravenous benzodiazepines must be avoided because the duration of their effect can be unpredictable and the patient may remain confused at the wake-up test. The key medication during this procedure is propofol (Diprivan, Stuart Pharmaceuticals, Wilmington, DE). This is an intravenous hypnotic agent which has a rapid onset of action when given intravenously and whose effects last only a few minutes. The best way of administering it is through intermittent boluses or continuous infusion. The patient usually wakes up promptly and lucidly when the drug is discontinued. Despite the fact that with increasing dosage propofol will eventually cause respiratory depression, a regimen can usually be found where the patient is adequately sedated without excessive respiratory depression. Constant interaction between the surgeon and the anesthesiologist is imperative.

Percutaneous electrode placement

Most commonly, the procedure is performed in the prone position. If the patient is unable to tolerate that position the procedure can also be performed in the lateral decubitus or seated position. Although feasible, the lateral decubitus position makes intraoperative fluoroscopic assessment more problematic. In some centers, the procedure is routinely performed with the patient in a sitting position. An advantage of this latter position is enhanced patient comfort and increased ability to obtain a thoracolumbar kyphosis, which facilitates insertion of the needle in the epidural space.

The patient is positioned in a comfortable prone position on a padded fluoroscopy table. A certain degree of kyphosis, as obtained by inserting a pillow underneath the abdomen, facilitates electrode insertion. It is very important to make sure that the trunk (for thoracolumbar placement) and/or neck (for cervical placement) are in a neutral position without any rotation.

Several considerations determine the level of electrode insertion. Since at least 3” of the lead body must lie within the epidural space in order to assure maximal stability of the electrode and minimize unwanted dislodgment, the insertion must be at least two spine segments below the desired target. For cervical placement, electrode insertion should be performed below the T1–2 level in order to avoid the risk of damaging with the needle the cervical cord enlargement. Most commonly, implantation for a pain problem in the lower part of the body usually entails electrode insertion at T12–L1 or L1–2 while implantation for an upper extremity target requires insertion at T2–3 or T3–4.

The fluoroscopy equipment is commonly utilized in both the anteroposterior (AP) and lateral planes at the time of needle insertion. This will allow monitoring of the depth of penetration and the laterality of the needle. The Tuohy needle is inserted with as shallow an angle as one possibly can obtain. While in the thoracic area this can be accomplished with either a midline or paramedian approach; in the upper lumbar area a paramedian approach is required. Besides lessening the risk of a dural puncture, a shallow trajectory greatly facilitates the subsequent insertion of the electrode in the epidural space. An excessively steep angle of entry of the electrode into the epidural space will also increase the risk of electrode fracture or dislodgment. A paramedian needle insertion is always more desirable, not only because it allows a more shallow angle of needle placement, but also because it avoids placing the electrode between two adjacent spinous processes. During lordotic extension of the trunk, the electrode can be pinched between the two spinous processes, leading to premature electrode fatigue and fracture.

A specific tactile feedback occurs when the needle engages the ventral surface of the lamina. When this is perceived, gentle wiggling of the needle hub will produce slight movement of the needle shaft but not of the needle tip. If entry in the interlaminar space proves to be extremely difficult, one can attempt to increase the degree of spine flexion to open up the interlaminar space; other options include changing the angle of approach or trying a different level. If the ligamentum flavum is calcified (as often happens in the thoracic spine), one can gently tap the needle to force it through; occasionally the only way to insert the needle through a calcified ligament is through an open laminotomy.

Of the several methods that are available to identify the epidural space, the tactile feedback that ones perceives when the tip of the needle enters the epidural space is important, but can be missed and cannot be completely relied upon. The most common method is the loss of resistance using a low-friction syringe (and not the usual disposable plastic syringes). After several needle insertions at one spine level, however, the loss of resistance method may loose its reliability. The insertion of a guide wire through the needle provides invaluable information as to the degree of penetration into the spinal canal. If the needle tip is in the interspinous ligament and has not completed penetrated the ligamentum flavum, the wire cannot be advanced. Advancement of the wire is possible only if the needle tip is in the paraspinal muscles or within the spinal canal. The pattern of advancement and the location of the wire under fluoroscopic imaging usually clarify its position.

Once the electrode is in the spinal canal, one has to be certain that it is positioned in the epidural and not subarachnoid space. Even though this might seem obvious and easily recognizable, there are instances in which it is more difficult. For example, when multiple attempts are needed to achieve seemingly proper needle placement, the ability to differentiate electrode location may be compromised. This is especially true if the arachnoid has been previously pierced and CSF has escaped and pooled in the dorsal epidural space. When the electrode is in the subarachnoid space, much less resistance is encountered when moving the guide wire (or the electrode), particularly for lateral movements. The wire seems almost to be ‘floating’ and undergoes large shifts of direction; this contrasts with epidural placement, where electrode movements are more discrete and obtained only with specific manipulations. The same type of wire/electrode movement can, however, be experienced epidurally if the dural sac has significantly collapsed because of CSF escape. Another helpful clue occurs during stimulation. When electrical current is delivered into the subarachnoid space, motor or sensory responses are elicited at a much lower threshold than that which occurs epidurally.

When the epidural space is satisfactorily identified, the electrode is gently inserted under fluoroscopic guidance in the AP plane. Removal of the electrode once it has been inserted through the tip of the needle has to be accomplished with the utmost care. It is very easy to catch the electrode in the needle and shear it at the junction of the insulation with the electrical contacts. If the electrode does not slide without any resistance (even minimal), the needle and the electrode should be removed altogether. Every time the electrode is withdrawn through the needle, it should be thoroughly inspected for minute breaks in the insulation, which would demand its disposal. Alternatively, a sleeve can be inserted over the guide wire in the epidural space. The guide wire is then removed and the electrode inserted through the sleeve. This obviates the risk of shearing the electrode during its manipulations. The electrode is then steered in the epidural space to the desired location. Should the targeted location prove to be less than two spinal segments from the electrode insertion, the electrode should be withdrawn and repositioned at a more caudal level.

Frequently, the electrode curves around the dural sac and slides in the ventral epidural space. In the AP projection this might be undistinguishable from a proper midline dorsal location. A gentle lateral curve of the electrode shortly after its entry in the epidural space should raise the suspicion that the electrode is actually going around the spinal sac. Absolute confirmation of the ventral location arises from the stimulation, which elicits mostly motor contractions. Alternatively, observation in the lateral plane readily discloses the anterior position of the electrode tip.

If more than one electrode is inserted, all the needles should be inserted before passing the electrodes into the epidural space. Needle insertion might shear an already implanted electrode. In addition, it is often possible to insert two electrodes simultaneously and advance them synchronously in the epidural space while maintaining their relative position and spacing. In other situations it might be convenient to insert a guide wire parallel to the electrode to block its passage to an unwanted location.

The electrode is then guided to the desired location. Some electrodes have a removable wire stylet that can be bent. Other electrodes are inherently stiffer and must be slightly bent to help in steering. Patience, persistence, and frequent use of the fluoroscopic unit are the keys to successful electrode placement.

Once in place, the electrode must be secured to the interspinous ligament to minimize dislodgment. Various anchors exist to facilitate this process.

Plate electrode placement

Patient positioning

Two basic positions can be utilized: prone or semilateral. The prone position allows a more straightforward understanding of the spatial relations and is one that the surgeon is usually more familiar with for spine surgery. However, the potential difficulty in airway management precludes any type of substantial intravenous sedation. This position is therefore contraindicated with cervical implantations under monitored anesthesia. The prone position can safely be utilized for implantation under general anesthesia.

In the semilateral position the patient lies comfortably in a bench-park type position, allowing the operative field to include the spine as well as the flank, abdomen or buttock for the pulse generator implant. If the pain is predominant on one side, the patient is asked to lie on the least affected side. In this position, airway management is safer than in the prone decubitus and the anesthesiologist feels more comfortable waking up and extubating the patient intraoperatively. One has to be aware that, because of variable degree of rotation of the body, 3-D spatial rapport and angles may vary; this can render 3-D visualization of the operated structures less intuitive. This problem is compounded in the cervical area since one has both rotation and flexion/extension of the spine.

In the author’s experience of more than 2000 implanted plate electrodes, the semilateral position has proven to be superior for implants in which the patient is awakened intraoperatively. The prone position is always utilized in the case of electrode placement at C1–2 (under general anesthesia).

Strategies at different spine levels

In the planning phase of the procedure, the implanting surgeon must be aware of the varying angulation of the spinous processes at the different spine levels. The surgeon must also be aware of the correlation between the various spine levels and the patterns of stimulation-induced paresthesiae.12,13

Thoracic – upper lumbar area

Prior to the surgery, an X-ray is taken in the operating room with metallic markers placed on the skin at the level of the planned incision. This allows for a precise marking of the entry level. Alternatively, one can localize the level with fluoroscopic imaging. In the lower thoracic – upper lumbar area the incision is usually placed over the interspinous gap. If placed over the spinous process it usually has to be lengthened. In the mid – upper thoracic area, instead, the incision can be placed either over the interspinous gap or over the spinous process itself (depending on the location of the incision, the bone removal technique may vary.). In a thin individual the incision is usually about 1 in length. Even in large individuals, the incision seldom needs to be more than 2 long. Different considerations apply if one is implanting in an area where a previous laminectomy has been performed.

Subperiosteal dissection is usually limited to the upper half of the spinous process inferior to the desired ligamentum flavum and to at least the inferior two-thirds of the spinous process superior to it. Parts of the superior spinous process are incrementally removed until the ligamentum flavum is visualized. In the lower thoracic – upper lumbar area this usually requires removal of inferior one-third of the spinous process. In the mid thoracic area the whole spinous process must be removed (due to the acute angle and significant overlapping of the spinous processes). A slightly different strategy can be adopted in the mid – upper thoracic spine. The incision can be made over the spinous process and removal encompasses the lower one-third of the superior process. Removal continues through the upper half of the inferior spinous process and lamina until the epidural space is exposed. With this approach the exposed epidural space lies directly under the incision instead of being slanted in a cephalad direction. This makes it possible to insert an electrode either in a caudal or cephalad direction. Following removal of the ligamentum flavum, the electrode(s) is inserted in the dorsal epidural space; the patient is then awakened and, when mentally clear, intraoperative testing begins.

Mid – low cervical area

The patient is placed in the semilateral position with the neck slightly flexed. The skin incision is usually about 1″ long. Even with a short skin incision one can reach 3–4 levels by extending the dissection of the subcutaneous tissues and by stretching the skin edges with a Gelpi retractor. The neck should be positioned so that it is flexed but not excessively rotated laterally. Even though some neck rotation is inevitable, extreme rotation increases substantially the difficulty of the procedure.

It is important to study in detail the anatomy of the spinous processes on the spine X-ray immediately before starting the procedure. Once inside the cervical fascia, one can palpate the spinous processes and try to recognize them based on the radiographic features. Because of the neck rotation, finding the midline in the cervical area is usually more difficult than in the thoracic area. The alignment may substantially vary even between two adjacent spinous processes and the plane from one spinous process to the next might be different; one could therefore be misled and wander off midline in the lateral regions of the neck. One has to be aware of the fact that, especially at C3–4 or C4–5, the spinous processes are small and the ligamentum flavum can be extremely thin. When the neck is flexed, one could inadvertently enter the spinal canal. It is not unusual in the cervical spine to be able to remove the ligamentum flavum and have adequate access to the epidural space without having to perform any bone removal.

C1–2

This placement is indicated in some patients with motor disorders or in pain management when the pain is in the jaw area and/or in the neck – shoulder – upper extremity(ies).

The patient is under general anesthesia and is placed in the prone position with the head held in a Mayfield head-holder (Fig. 30.8). The cervico-occipital junction is flexed as much as possible to open the C1–occiput space but the cervicothoracic junction is kept straight without angulation.

An incision of 1–1.5″ is placed at the cervico-occipital junction, over the arch of C1. After exposing the arch of C1 and the superior aspect of the C2 lamina, the ligamentum flavum between C1 and C2 is removed. Complete subperiosteal dissection is then carried out underneath the arch. If the C1 arch is particularly thick, it is undermined and thinned with a 45° Kerrison rongeur. The electrode(s) is then passed in a caudal direction under the arch of C1 and then under the C2 lamina. No electrical contact should lie above the arch of C1, in correspondence of the cisterna magna. At this level the distance between the electrode and the cord makes it almost impossible to stimulate the spinal cord. Intraoperative testing is carried out relying on motor responses. Stimulation is applied at 1–2 Hz with increasing voltage until motor contractions are elicited. At this level a midline electrode usually triggers contractions in the neck and shoulder muscles bilaterally. A more laterally placed electrode produces motor stimulation in the ipsilateral upper extremity. By sliding the electrode laterally one can precisely fine-tune the degree of lateralization of the stimulation. At this level, sensory stimulation usually follows closely the motor stimulation.

Implantation of the radio receiver/pulse generator

There are two types of spinal cord stimulation systems: (1) those with three parts, namely the electrode, an extension cable and the radio receiver/pulse generator, and (2) those with two parts only, namely the electrode and the radio receiver. In this latter situation, the electrode extension is long enough to connect directly to the radio receiver/pulse generator. The position of the pulse generator/receiver is crucial. If placed in an uncomfortable position, the unit will eventually have to be moved to a different location. The most common placement sites include the abdomen, the fat pad in the posterior iliac area, the infraclavicular area and the lateral chest wall area (Fig. 30.9).

Infraclavicular area

This location can be utilized exclusively with cervically implanted electrodes. It is an excellent location for both RF and lithium-powered systems. The presence of the underlying rib cage facilitates contact between the antenna and the radio receiver. A potential problem with this location, particularly in females, is that, in some instances, the unit tends to migrate caudally and lodge in the breast area. Even securing the unit to the subcutaneous tissues with nonreabsorbable sutures may not provide a solution. The author has actually secured the unit to the bone by drilling a whole through the clavicle and suturing the unit to it. The subcutaneous pocket is made with either blunt (fingers, Kelly clamps) or sharp (scissors, cutting coagulator) dissection. When preparing the pocket for an RF system, one must make sure that the thickness of the subcutaneous tissue above the receiver is not more than 0.5–1″. A thicker layer might impede transmission of the RF signals from the antenna to the receiver and prevent activation of the system. The pocket must be only slighter larger than the implanted unit in order to prevent its migration. Attention must be paid to having the whole unit (including the connectors) completely within the pocket and not crossing underneath the skin incision. The excess cable should be looped behind the unit, so that it would not be damaged if the subcutaneous pocket were to be tapped with a needle to aspirate fluid. The unit can be secured to the underlying tissues with nonreabsorbable sutures. If one is concerned about a high risk of migration or tilting of the unit (in patients with large adipose tissue and/or loose subcutaneous tissues), the unit should be placed in a Dacron pouch which can then be secured with multiple sutures to both the deep and superficial subcutaneous layers. This maneuver will minimize, but not completely eliminate, the risk of migration/tilting.

Placement under the fascia is sometimes indicated, particularly in thin patients or in subjects who cannot tolerate the bulging of the implant under the skin. The subfascial plane is often vascular, because of the vascularization of the underlying muscles, and adequate hemostatis must be secured prior to closure. Careful securing of the implant is also mandatory to prevent migration of the unit in the subfascial plane.

Intraoperative testing

If the procedure is performed under monitored or local anesthesia, the patient is woken up after the electrode(s) is placed in the desired position. The patient must be fully awake, cooperative, and with normal cognitive abilities before starting the intraoperative testing. Failure in this respect will inevitably result in incorrect positioning of the electrode. It is extremely important to assure that the patient is comfortable during the stimulation trial. When testing, the implanter must be aware of the different responses that stimulation of the various intraspinal structures will elicit. The strategy for electrode placement in both the transverse and longitudinal direction varies according to the pain topography and intraoperative elicited responses.12 As a general rule, one wants to achieve a complete paresthesia coverage of the painful area at the lowest possible stimulation threshold and with the least amount of extraneous stimulation.

If the implantation is performed under general anesthesia, one can rely on three factors.

References

1 Diascorides T. Goodyear J., editor. The Greek herbal of Diascorides. London: Oxford University Press, 1934.

2 Galen C. Tallmadge M., editor. De usu partium. Ithaca, NY: Cornell University Press, 1968.

3 Kellaway P. The part played by electric fish in the early history of bioelectricity and electrotherapy. Bull Hist Med. 1946;20:112-137.

4 Stillings D. A survey of the history of electrical stimulation for pain to 1900. Med Instrum. 1975;9:255-259.

5 Shealy CN, Mortimer JT, Reswick JB. Electrical inhibition of pain by stimulation of the dorsal columns. Preliminary clinical report. Anesth Anal Cur Res. 1967;46:489-491.

6 Hoppenstein R. Electrical stimulation of the ventral and dorsal columns of the spinal cord for relief of chronic intractable pain. Surg Neurol. 1975;4:187-194.

7 Larson SJ, Sances A, Cusick JF, et al. A comparison between anterior and posterior implant systems. Surg Neurol. 1975;4:180-186.

8 Lazorthes Y, Verdie JCL, Arbus L. Stimulation analgesique medullaire anterieure et posterieure par technique d’implantation percutanee. Acta Neurochir. 1978;40:253-276.

9 Waltz JM, Andreesen WH. Computerized percutaneous multi-level spinal cord stimulation in motor disorders. Appl Neurophys. 1982;45:73-92.

10 North RB, Kidd DH. A prospective randomized comparison of spinal cord stimulation electrode design. Acta of the 8th World Congress, The Pain Clinic. Tenerife, Spain, May 6–8 1998:p. 57.

11 Barolat G, Zeme S, Ketcik B. Multifactorial analysis of epidural spinal cord stimulation. Stereotactic Functional Neurosurg. 1991;56:77-103.

12 Barolat G, et al. Mapping of sensory responses to epidural stimulation of the intraspinal neural structures in man. J. Neurosurg. 1993;78:233-239.

13 Barolat G. Experience with 509 plate-electrodes implanted epidurally from C1 to L1. Stereotactic Functional Neurosurg. 1993;61:60-79.