Chapter 74. Cryoanalgesia and Radiofrequency

Introduction

Cryoanalgesia, that is, the use of low temperatures to provide analgesia, is a pain management technique that can be applied to a variety of painful situations. From the topical application of ice to a sprained ankle or cryoneurolysis of an intercostal nerve for postoperative pain control in a post-thoracotomy patient to cryoneurolysis of the trigeminal nerve for trigeminal neuralgia, cryoanalgesia has a multitude of uses. This chapter focuses on the application of cryoanalgesia for acute and chronic pain management.

Historical Perspectives

The first account of the use of cryoanalgesia for pain control was recorded by Hippocrates (460–377 bc). He described the use of ice and snow packs to prevent pain from surgery.1 An 11th century Anglo-Saxon monk, Avicenna of Persia (980–1070), and Severino of Naples (1580–1656) also recognized the use of cold as preemptive analgesia for surgery.1,2 Other historical recorders who recognized the value of cryoanalgesia include Napoleon’s surgeon general, Baron Larre (1812), Dr James Arnott (1851), and Richardson (1866).1,2

Recently, Irvine Cooper and his colleagues (1961) developed the first prototype of a cryoprobe that could be used for analgesia.1,3 It utilized the principle of phase change using liquid nitrogen to produce a temperature of –196°C. Several years later in 1967, Amoils developed a smaller cryoprobe that employed the Joule-Thompson principle and used carbon dioxide or nitrous oxide to produce temperatures to –50°C.4) Present cryoprobes use either of these two principles to produce a temperature of at least –70°C.

Cryoneurolysis involves using cold to induce nerve injury, thus providing analgesia. The exact cellular mechanism of nerve injury is not known. Theories include ischemic necrosis, physical destruction by large cellular ice crystals, damage to proteins, minimal cell volume, production of autoantibodies, and membrane rupture caused by rapid water loss.5 Cryolesioning causes a second-degree nerve injury as classified by Sunderland. A second-degree injury involves degeneration of the axon and myelin sheaths (wallerian degeneration) from the site of freezing distally to the nerve’s termination.6 The minimum temperature required is 20°C or lower.5 The endoneurium, perineurium, and ectoneurium remain intact, thus regeneration is possible. The return of normal sensory and motor activity depends primarily on two factors: the rate of axonal regrowth (average of 1 to 2 mm per day) and the distance of the cryolesion from the end organ, although nerve conduction velocity remains only reduced to an average of 35 days.1,7,8

The success of a cryolesion depends on several factors: the rate of freezing and thawing, the temperature attained by the tissue in proximity to the cryoprobe, and the size of the cryolesion.7 Evans and Gill et al showed that repeat freeze-and-thaw cycles enlarged the size of the cryolesion, improving the success of the block.5,9 Freeze cycles are typically 1.5 to 3 minutes with a 30-second thaw period. Evans et al, using exposed rat sciatic nerves, showed that a minimum temperature of –20°C was necessary for prolonged sensory loss and that further decreases in temperature did not greatly influence the duration of sensory loss.7 The size of the cryolesion is also important since saltatory conduction is still possible in myelinated fibers, where there are short interruptions in axonal continuity. Douglas and Malcolm in 1955 demonstrated in cats that an injury of 3 to 6 mm in length was sufficient to prevent saltatory conduction.10

The modern day cryoprobe creates low temperatures either by the expansion of a compressed gas (Joule-Thompson effect) such as nitrous oxide or carbon dioxide, or by the principle of phase change with liquid nitrogen (Fig. 74-1A and 74-1B). The expansion of a compressed gas through a small orifice can generate temperatures of –70° to –80°C and the gas most commonly used is nitrous oxide.1,3 Liquid nitrogen cryoprobes can reach temperatures of –196°C.

The cryoprobes come in a variety of sizes and shapes. Most have a Teflon coating for insulation, and incorporate a stimulator to assist in locating the intended nerve and a thermocouple to measure core temperature. The stimulator and the thermocouple are located in the tip of the cryoprobe. The size of the freeze zone for a gas expansion cryoprobe is two to three times the probe’s diameter, whereas liquid nitrogen cryoprobes create a lesion that is three to five times the size of the probe. The typical size of a cryoprobe ranges from 12 to 18 gauge (Fig. 74-2).

Cryoneurolysis is best suited for conditions in which the nerve to be lesioned is small and well localized. It can be done by a surgical incision, as seen with post-herniorrhaphy and post-thoracotomy pain, or closed, that is, percutaneously. The percutaneous approach is usually done for neuroma and nerve-entrapment syndromes. Specific lesions for acute and chronic pain are discussed later in this section. Contraindications include patient refusal, lesioning of a motor nerve, and the use of cryoneurolysis when a diagnostic nerve block has failed.

Postoperative Pain Management

Cryoneurolysis is beneficial in a variety of surgical procedures to provide postoperative analgesia. Unfortunately, clinical studies have only looked at a few such applications. Herniorrhaphy and thoracotomy are the two most common surgeries in which cryoneurolysis has been used for postoperative pain management.

Post-Herniorrhaphy Pain

Herniorrhaphy provides an ideal situation in which a cryolesion could be used to provide postoperative analgesia. The ilioinguinal nerve is usually identified and retracted during the initial dissection of a herniorrhaphy. Wood et al in 1981 showed that a cryolesion of the ilioinguinal nerve after herniorrhaphy significantly decreased the amount of oral analgesics required for postoperative pain control.11 This study compared cryoneurolysis with oral analgesics alone and paravertebral block for postoperative pain control in patients undergoing day-case herniorrhaphy patients. Prior to closure of the herniorrhaphy incision, the ilioinguinal nerve was elevated and frozen until a visual lesion was noted. Not only did the patients in the cryoanalgesia group require less oral medications but they also had improved appetites, were more active, and returned to work sooner. Even though this was only a small study, it showed the benefits of cryoneurolysis in allowing patients to have less pain and enabling them to return to work earlier.

Post-Thoracotomy Pain

Analgesia following thoracotomy is particularly important not only to keep the patient comfortable but also to facilitate extubation and prevent respiratory sequela via adequate pulmonary toilet. Oral analgesics have the propensity to diminish the cough reflex and to decrease respirations in the doses necessary to adequately control postoperative pain.12 Intercoastal nerve blocks with local anesthetic generally do not provide effective long-term analgesia and frequently have to be repeated. In several studies, intraoperative cryoneurolysis of the intercostal nerves prior to closure of the thoracotomy incision has been shown to be effective in decreasing postoperative pain and the amount of oral and parenteral analgesics required post-procedure.12–14 Before closure of the thoracotomy incision, the intercostal nerve at the site of the incision as well as the nerves two levels above and below the incision are dissected away from the intercostal vessels and cryoablated until a visual lesion is observed.

Chronic Pain Management with Cryoneurolysis

A multitude of conditions exist in the realm of chronic pain management in which percutaneous cryoneurolysis can be applied. Of note is that it should only be performed after a successful diagnostic nerve block with local anesthetic. The following section describes specific instances in which cryoneurolysis has been effective for control of chronic pain.

Facial Pain

Goss in 1988 described the use of cryoneurotomy for the treatment of intractable temporomandibular joint pain.15 In this study, six female patients who had severe temporomandibular joint pain and had failed conservative treatment with oral therapy and surgery by others underwent open cryoneurotomy of the lateral and posterior joint capsule and open cryoneurolysis of the great auricular nerve. Five of the six were pain-free for over a year. Four of the six had recurrent pain and were successfully treated with repeat cryoneurotomy. Zakrzewska and Nally studied the role of cryoneurolysis in the management of paroxysmal trigeminal neuralgia16; 145 patients were treated and followed up from 1 month up to 6 years. The nerves cryoablated included the infraorbital, mental, long buccal, lingual, greater palatine, and posterior-superior dental. Fifty-eight percent of 89 infra-orbital nerves treated were pain-free at 1 year, 36% of 85 mental nerves treated were pain-free after 1 year, and 66% of 23 buccal nerves treated were without pain after 1 year. Barnard and colleagues studied 21 patients with intractable facial pain, including postherpetic neuralgia, atypical facial pain, post-traumatic neuralgia, and facial pain secondary to malignant disease.2 After cryoablation, the postherpetic neuralgia group had a median duration of pain relief of 38 days and ranged from 0 to 84 days. The other types of facial pains were grouped together and had a median duration of pain relief of 116 days and ranged from zero to 257 days. These studies indicate that cryoneurolysis has a place in the management of intractable facial pain.

Thoracic Pain

Common causes of chronic thoracic pain include intercostal neuralgia secondary to a thoracotomy scar, postherpetic neuralgia, nerve root pain due to vertebral collapse, and carcinoma. Cryoneurolysis of the intercostal nerve has been shown effective in alleviating pain due to intercostal neuralgia. Jones and Murrin did a retrospective study of 70 patients from 1982 to 1984 who had intercostal pain from several causes, namely scar pain from thoracotomy and postherpetic neuralgia.17 They found that cryoneurolysis was more effective for control of intercostal pain for patients with pain secondary to thoracotomy scar than for postherpetic neuralgia pain. This procedure is only performed when a diagnostic intercostal block with local anesthetic has given effective analgesia. Cryoneurolysis is performed percutaneously at the angle of the rib. Using aseptic technique, the skin is infiltrated with local anesthetic and an intravenous cannula (usually 12 to 14 gauge) is inserted at a 45 degree angle to the skin. The cannula is aimed at the inferior border of the rib. This can be done by palpation or under fluoroscopic guidance. The cryoprobe is then inserted through the cannula and walked under the rib. Sensory stimulation can be used to reproduce the pain. Two lesions are performed for 90 seconds, each with 30-second thaw periods. A post-procedure chest radiograph should be obtained to rule out pneumothorax.

Low Back Pain of Facet Origin

Facet rhizotomy has been performed with percutaneous cryoneurolysis in patients with a facet origin of low back pain. The goal is to lesion the common branch of the posterior primary ramus as it crosses the medial portion of the transverse process near the superior pars of the facet joint. Since the facet joint is innervated by nerves from multiple levels, this procedure must also be performed one level above and one level below the facet joint in question. A positive diagnostic common branch nerve block must be obtained prior to performing this procedure. Forty-seven of 50 patients in a study done by Schuster received significant pain relief after facet rhizotomy with cryoneurolysis.18

Peripheral Nerve Pain

Cryoneurolysis of peripheral nerves can be performed in cases of nerve entrapment or painful neuroma. Only pure sensory nerves or nerves with only a minor motor component should be cryolesioned. Precise nerve mapping should be done prior to cryoneurolysis to prevent adverse sequelae, such as lesioning a motor nerve. This can be done with diagnostic nerve blocks with local anesthetic.

Perineal Pain

Entrapment of the ilioinguinal nerve with resulting neuralgia is occasionaly seen in post-herniorrhaphy patients. A diagnostic block should be performed to rule in or rule out the diagnosis. If correctly diagnosed, a percutaneous cryolesion can be performed. Racz and Hagstrom studied 25 patients with iliohypogastric and ilioinguinal nerve entrapment secondary to a variety of causes.19 After cryolesioning, 46% reported good to excellent pain relief.

Cryoneurolysis of sacral nerve roots has been performed for cancer pain, coccydynia, and sciatic nerve pain.3 Sciatic pain can be partially treated with S1 and S2 cryolesions, whereas bilateral S4 cryoneurolysis can help alleviate pain secondary to coccydynia and rectal cancer. Loev et al reported a novel approach to cryoablation of the ganglion impar in a 73-year-old patient with chronic anal and perineal pain secondary to surgical resection of carcinoma.20 He and his colleagues cryoablated the ganglion impar by passing the cryoprobe through the sacrococcygeal ligament under fluorosopic guidance. The patient had 80% pain relief for 6 weeks, at which time it was repeated secondary to returning pain.

Most complications from cryoneurolysis are secondary to improper technique. If an introducer sheath is not used, frostbite can occur at the skin-entry site and cause ulceration, especially if superficial nerves, such as the supraorbital or infraorbital nerves, are cryoablated. Motor nerves can be inadvertantly lesioned if proper localization of the intended nerve is not done. Cryoneurolysis is not a permanent process and motor function will return over time. Although infrequent, there have been reports of dysesthesia following cryoneurolysis of intercostal nerves.20

Cryoneurolysis is a technique of prolonging analgesia without serious sequelae. It has many applications ranging from postoperative pain control to alleviation of many chronic pain states. It is a relatively safe and effective method of pain control. Data on its efficacy is however poor. Further detailed data will have to be obtained and compared with other neuroablative techniques prior to identifying it as a reliable analgesic technique.

Introduction

Interventional pain physicians continue to strive to find techniques that provide relief to chronic pain patients. The techniques sought must be percutaneous, discrete, reliable, reproducible, effective, and most of all, safe. Though not a new concept, radiofrequency as a method to interrupt painful pathways continues to be refined in its technique and broadened in its application. It provides a useful tool for the pain physician to provide long-term pain relief to patients in a safe, effective, and reproducible manner.

History

The use of radiofrequency energy for the creation of lesions may be traced to as early as 1863 when Beaunis created a direct-current lesion in the animal brain.21 Fournie followed with the development of bipolar lesions in the animal neural tissue.21 In the mid-1920s, Harvey Cushing did some basic research utilizing the potential of radiofrequency power for electrosurgery.22,23 The first known use of radiofrequency lesioning for the treatment of chronic pain was in 1931 when Kirschner successfully treated trigeminal neuralgia with the thermocoagulation of the gasserian ganglion.24,21 In 1953, William H. Sweet and Vernon Mark showed that the use of radiofrequency (RF) current for lesions production has decisive advantages over the time-tested direct current lesions that had been used for decades.25,26 They demonstrated that the RF lesion is characterized by well-circumscribed borders that have better control of the size and shape compared with the lesion from the direct-current method.25,26 In the late 1950s, the work of Cosman and Aronow made the first commercial RF lesion generator available.27,21 In 1960, Mundinger et al28 emphasized the importance of monitoring the electrode temperature in RF lesion making to obtain consistent lesion size.28,29 In 1974, Sweet and Wepsic30 performed percutaneous retrogasserian thermal rhizotomy on 274 patients with facial pain (214 of whom had trigeminal neuralgia). Noting that touch was preserved in some or all of a trigeminal zone rendered analgesic, the authors concluded the heavily myelinated Aβ fibers were more resistant to heating.30 This was consistent with the findings of Letcher and Goldring in 1968.31 Their study demonstrated in vitro the effects of RF and heat on the peripheral nerve action potential in the cat. Their demonstration that RF heat lesions block the action potentials of smaller delta and C fibers before those of the alpha-beta group of fibers formed the neurophysiologic basis for the therapeutic use of RF lesions.31,32 Uematsu,33 however, in 1977 described a preliminary histologic study that demonstrated the effects of RF thermocoagulation on the sciatic nerve of the cat. This study found indiscriminate destruction of both small and large myelinated fibers rather than selective destruction of the smaller myelinated fibers.32,33 This was supported in 1981 by Smith et al32 who studied RF lesions at 85°C in dogs. This study noted that RF lesions were not specific for unmyelinated C fibers, small myelinated fibers, or large myelinated fibers. They also demonstrated the lack of specificity for graded lesions of 45°, 55°, 65°, and 75°C.32 Tew34 significantly advanced the technique using a spring-loaded curved radiofrequency electrode tip that could be precisely stereotactically placed to lesion the first, second, or third branch of the trigeminal nerve.35 The first use of RF lesioning for chronic spinal-based pain came in 1975 when Shealy performed posterior primary rami radiofrequency in the lumbar region.36,21 In the 1980s, Sluijter advanced the applications of RF for spinal pain by refining percutaneous techniques for cervical, thoracic, lumbar, and sacral spine syndromes.35 Sluijter and Mehta modified the existing electrode for RF, making it smaller and capable of temperature monitoring.37 This allowed the passing of current through electrodes to reproduce the patient’s pain with small-diameter RF needles. These needles were placed using local anesthesia with fluoroscopic guidance. After reproduction of the pain, a local anesthetic was placed on the targeted structure followed by painless lesioning.35

There are many techniques for local destruction of nervous tissue in the brain, spinal cord, or peripheral nervous system. The methods commonly used by interventional pain management specialists include38:

• 1. Chemical injection (chemical neurolysis)
• 2. Cryoneurolysis (freezing of the nerve)
• 3. Lasers
• 4. Radiofrequency thermocoagulation.

Chemical injection of a neurolytic such as alcohol, glycerin, or phenol can be advantageous for disruption in an area with extensive, weblike innervation, as in the sympathetic nervous system. These lesions, however, are irregular, variable in shape, and the spread can be impossible to control. This can lead to problems with reproducing an effective block at a later date. A postneurolytic neuritis from aberrant regeneration of nerve endings can be a serious, unwanted side effect in patients treated with chronic, nonmalignant pain. Cryoneurolysis is an incomplete destruction that avoids the side effect of deafferentation pain caused by RF in peripheral nerves. Therefore, in peripheral nerves, cryoneurolysis may be more appropriate than RF. However, the size of the probes (diameter >3 mm) can be of concern and may limit the spectrum of applications. The duration of effect is short: only a matter of weeks or a few months. Lasers may be used for lesioning; however, it is difficult to quantify the extent and rapidity of destruction. Also, because of the inadequate temperature monitoring, there is variability of effect, depending on tissue parameters such as blood flow and thermoconductivity.38

Radiofrequency thermocoagulation offers several advantages when compared to other methods of interrupting nerve pathways. The lesions from a RF needle have smooth, well-circumscribed borders. The lesion sizes may be reproducibly quantified by monitoring the temperature of the tip of the electrode. Monitoring of the temperature diminishes the risks of unwanted side effects such as boiling, charring, and sticking.29 The target for lesioning is localized by monitoring stimulation and impedance. RF needles have different lengths (e.g., 5, 10, and 15 cm) and electrode configurations that provide suitability for different anatomic sites.38

Radiofrequency is a high-frequency current that is applied to neural tissue through a closed circuit. The current, produced by a generator, runs in the active electrode (the needle, or where the lesion is made) and disperses out the dispersive electrode. The difference in the currents between the active electrode and the dispersive electrode is the voltage, which is monitored. The body tissue completes the circuit. The voltage generates an electric field that runs between the active (lesion) and the dispersive electrodes. This electric field causes electric force on the ions in the tissue (electrolytes move back and forth). The tissue then heats the electrode tip that in turn absorbs heat from the tissue. Therefore, an equilibration of the temperature between the tissue and the electrode tip occurs. The electrode tip temperature represents the temperature of the hottest tissue adjacent to it.38

The lesion size is determined by four major factors: (1) electrode size and configuration, (2) the rate of thermal equilibrium, (3) the temperature generated, and (4) local tissue characteristics.39 The lesion size may be altered by the distance from the active electrode tip and the electrode diameter. The lesion is spheroidal around the active electrode and tapers at the tip so that minimal lesion formation occurs beyond the tip of the active electrode. In clinical practice, the placement of the active electrode should be oriented in an axis that will produce a spheroidal lesion along the desired location for neurolysis.40,41 The lesion length is 0 to 2 mm longer than the exposed electrode tip length. The lesion diameter rises exponentially but plateaus at 30 seconds, though thermal equilibrium is reached at 60 seconds.29 It can provide a 4- to 5-mm circumferential lesion around the radius of the length of the electrode. For example, an electrode 3 mm in length and 0.4 mm in diameter will form a lesion 5.5-mm long.38 According to Cosman,29 the lesion will extend 1.5 electrode diameters beyond the tip distance. For maximum lesion area, the electrode must be placed parallel to the nerve. The larger and more complete the lesion is, the longer the expected pain relief. Over time, the nerves regenerate and the pain will return. Therefore, a larger electrode will generate a larger lesion. However, it will also cause more tissue destruction on insertion, unwanted neural destruction, and larger reversible zones.29

Temperature monitoring creates more consistent, safer lesions. Brodkey et al established that irreversible nerve blocks occur at 45°C in the brain.42 It is also known that at >90°C the tissue will boil, causing adherence to the probe and tearing of the tissues. Therefore, a tip temperature of >100°C must be avoided.29 At 95°C, one might notice steam formation, charring, sticking and increased risk of hemorrhage as adherent tissues tear.29 On the RF machine, a sudden resistance may be noted followed by a drop in current flow to zero. At these high temperatures, one may see a “popping” lesion. A rapid decrease in the current meter and an increase in the voltage meter. Gas and steam at the tip may be forced up the shaft. The electrode gives a “popping” sound. This lesion produces ragged and unpredictable damage.29 Another advantage to monitoring the temperature is assurance that a lesion has been made.

Impedance monitoring gives independent information of the lesion process and can signal equipment failure.43 Pre- and post-impedance changes, at a minimum, tell the operator a lesion has been made, that is, irreversible structural changes have occurred. Impedance monitoring can increase the safety of doing an RF procedure because it is an instant continuity check.43 The RF circuit has numerous connections and wires that are vulnerable to unpredictable breakage or intermittency. The impedance monitor can alert the operator to an open-circuit problem, as well as the equally troublesome short-circuit situation.43

Unpredictable factors can alter the lesion size and shape. A large blood vessel near the electrode tip can serve as a heat sink, causing local cooling and irregularity of the lesion shape.29 Cerebrospinal fluid may also act as a heat sink, drawing away the heat, and alter the lesion shape.29

On the histologic level, after the lesion is made, the appearance of the lesion is one of a local burn.39 This is a neuroablative procedure. After creation of the lesion, wallerian degeneration becomes apparent. The perineurium may also be destroyed.39 This makes RF more susceptible to neuroma formation than when cryoneurolysis is used. Though controversial, some argue there is selectivity for Aδ and C fibers with sparing of the larger Aβ fibers at lower temperatures.39 More investigation regarding this controversy is needed.

Accurate placement of the active electrode to ensure accuracy of the lesion may be done using fluoroscopy and electrostimulation. Biplanar fluoroscopy is used to target the nerve roots, facets, disks, and sympathetic nerves. Further specificity is obtained using water-soluble, nonionic contrast dye. Electrostimulation for sensory is performed at 50 Hz. Optimal placement of the active electrode near the dorsal root ganglion is considered if there is reproduction of the patient’s concordant pain at 0.4 to 0.7 V. If there is reproduction at less than 0.2 V, this may be suggestive of intraneural placement. This places the patient at greater risk for neuritis. Motor stimulation should also be tested at 2 Hz and 2 V. Fasciculation or titanic stimulation with this indicates placement near a ventral nerve root.

Before deciding on RF as part of the patient’s treatment algorithm, all conservative therapy must have been fully explored. This includes a thorough physical examination with appropriate diagnostic tests such as x-ray, magnetic resonance imaging (MRI), computed tomography (CT), and myelographic studies. The patient should be an active participant in a multidisciplinary pain management program that includes medical, psychological, and physical therapy evaluation and treatment. The pain should be chronic (>3 months). There should be no indication for surgical intervention and no invasive lesion at the site such as tumor or infection. A successful diagnostic block should be performed first. A prognostic block is not without its pitfalls, however. There is always concern for the placebo effect or lack of specificity of the block. Exclusion criteria includes patients with coagulopathies, psychopathology, ongoing septicemia, or neuropathic/deafferentation pain syndromes.

For RF to remain safe, precise, and be able to yield reproducible results, the equipment must be able to23,41:

• 1. Measure impedance.
• 2. Stimulate a wide range of frequencies.
• 3. Accurately time the duration of the lesion for a precise temperature measurement in the core of the lesion tissue.
• 4. Accurately measure and indicate amperage and voltage.
• 5. Gradually increase temperature with time.

It is also imperative to check the insulation coating of the RF needle for any imperfections as “cracks” that may lead to a leakage of current into tissue and unwanted injury. Patients with a pacemaker or spinal cord stimulator must also be monitored because there can be interaction of the RF equipment.41 As a consequence of RF lesioning, patients who have sensing pacemakers may have a period of asystole unless their pacemakers have an override. Be knowledgeable of the features of the pacemaker and use continuous electrocardiographic monitoring as indicated during procedures.41 Though not widely documented, the possibility of current passing from the active electrode of the RF needle in the direction of the spinal cord stimulator and involving the spinal cord may exist.41 Therefore, be sure the all spinal cord stimulators are turned off prior to initiation of the RF procedure.

Stellate Ganglion

The stellate ganglion is formed by the fusion of the inferior cervical ganglion with the first thoracic ganglion. This usually lies in the front of the neck of the first rib and extends to the interspace between C7 and T1. This contains the preganglionic axons and post-ganglionic cell bodies T1 to T9. Indications for blockade of the stellate ganglion are chronic pain conditions of the head, neck, and upper extremity:

• 1. Complex regional pain syndrome type I (reflex sympathetic dystrophy)
• 2. Complex regional pain syndrome type II (causalgia)
• 3. Herpes zoster
• 4. Postherpetic neuralgia
• 5. Phantom limb pain
• 6. Hyperhidrosis
• 7. Vascular headaches
• 8. Vascular insufficiency

Complications for RF of the stellate ganglion include:

• 1. Intraspinal injury
• 2. Intravascular injury
• 3. Pneumothorax
• 4. Recurrent laryngeal nerve paralysis
• 5. Phrenic nerve paralysis
• 6. Brachial plexus injury.

Radiofrequency of the stellate ganglion block should be performed under fluoroscopic guidance after a successful diagnostic block using local anesthetic and steroid.

The patient is placed in the supine position with a pillow under his or her shoulders to extend the cervical spine. Under direct anterior-posterior (AP) fluoroscopy, the C7 vertebral body is identified. A skin wheal is raised over the ventrolateral aspect of the body of C7 with 1 mL of local anesthetic and a 25-gauge needle.21 A 16-gauge angiocatheter is inserted through the skin wheal to contact the body of C7 in the ventrolateral aspect. This is at the junction of the transverse process and the vertebral body. Depth and direction should be confirmed in both the AP and lateral views frequently throughout the procedure. The needle tip will be positioned deep to the anterior longitudinal ligament. The longis colli will be lateral to the needle tip.44 A 20-gauge, curved, blunt cannula with a 5-mm active tip is guided through the angiocatheter. The tip should rest at the junction of the transverse process and the vertebral body. Further confirmation of proper placement of the needle is with the injection of a water-soluble, nonionic contrast medium. A sensory and motor stimulation trial must be performed because of the location of the phrenic nerve (lateral) and the recurrent laryngeal nerve (anterior and medial) to the lesion site.21 A sensory trial is performed at 50 Hz and 0.9 V. A motor trial is performed at 2 Hz and 2 V. While motor stimulation is performed, the patient should phonate, saying “ee” to assure preserved motor function. If there is disruption of the phonation, the needle tip must be repositioned. A small volume of local anesthetic (0.5 mL) should be injected prior to lesioning. The lesion is instituted for 60 seconds at 80° C. The cannula is then redirected to the medial most aspect of the transverse process in the same plane.21 Confirmation of placement in the ventral aspect must be demonstrated in the lateral view. Before lesioning, the patient must be retested for sensory and motor stimulation. A repeat dose of the local anesthetic should also be given through the cannula. A third and final lesion should be directed at the upper portion of the junction of the transverse process and the body of C7.21,44

Further sympathetic interruption of the upper extremity may be performed with a lesion at the T2 or T3 level. Sympathetic disruption at the thoracic level is indicated for the treatment of complex regional pain syndrome I (reflex sympathetic dystrophy), complex regional pain syndrome II (causalgia), arterial occlusions leading to ischemia, drug-resistant Raynaud’s disease, Buerger’s disease, and frost injuries of the upper extremities.47 As in the stellate ganglion lesion, a diagnostic block should be performed prior to any ablative procedure. Before performing the procedure, because of the relationship of the sympathetic chain at this level and large vascular structures, a coagulopathy must be ruled out (prothrombin time, partial thromboplastin time, and bleeding time within normal limits). The patient is placed in the prone position on the fluoroscopy table. The camera is placed in the oblique view (30 degrees) to place the shadow of the body lateral to the lamina (Fig. 74-3A). A cephalocaudad view is necessary to correct for the upper thoracic kyphosis (approximately 20 degrees). A 10- to 15-cm, blunt, curved cannula with a 10-mm active tip is introduced through a 16-gauge angiocatheter that has already pierced the skin approximately 3.5 cm from the midline. A tunnel view should demonstrate the cannula advancing toward the margin of the lamina. It is advanced along the lateral body until the tip is halfway between the anterior and the posterior borders of the body (see Fig. 74-4C). Proper needle placement must be assessed with the spread of the contrast medium (water-soluble, nonionic). This should demonstrate paravertebral spread under the pedicles in the AP view without intravascular, pleural, epidural, or intrathecal uptake (see Fig. 74-4B). Sensory stimulation must be performed at 50 Hz and 0.9 V. Greater than 0.9 V will produce a deep ache. Motor stimulation at 2 Hz and 2 V must be negative. Additionally, 1 mL of a local anesthetic and steroid mix (a 50:50 mix of 0.2% ropivacaine, 2% lidocaine, and 1 mL of 40 mg/mL triamcinolone) is injected after negative aspiration. A 30-second delay should be observed before lesioning. Two lesions are performed at each level with the active tip turned cephalad for the first lesion and caudad for the second. The lesions should be made at 80°C for 90 seconds. A postoperative end-expiratory chest radiograph is taken to rule out a pneumothorax. Unfortunately, 10% to 15% of patients will suffer a post-procedure neuritis. This can last for 3 to 6 weeks.44,46,47

Celiac-Splanchnic Plexus

The celiac plexus lies anterior to the aorta and crux of the diaphragm at T12. It is composed of efferent preganglionic splanchnic nerves, efferent sympathetic preganglionic nerves, and afferent nerves to the viscera of the upper abdomen. This supplies visceral innervation to the pancreas, liver, gallbladder, omentum, mesentery, and alimentary tract from the stomach to the transverse colon. The RF lesioning of the splanchnic nerves prior to joining the celiac plexus allows a discrete, precise lesion from a posterior approach that avoids the risk of puncturing the aorta. The indication for a lesion of the splanchnicnerves is pain originating from the viscera innervated by the celiac plexus. Complications associated with these lesions are:

• 1. Hypotension
• 2. Paresthesia of lumbar somatic nerve
• 3. Vascular injury (including thrombosis or embolism)
• 4. Intrathecal or epidural injury
• 5. Paraplegia
• 6. Pneumothorax
• 7. Chylothorax
• 8. Intradiscal placement of the needle
• 9. Retroperitoneal hematoma
• 10. Pain during and after procedure.48

A diagnostic block should precede the performance of RF. The patient should be placed in the prone position on a fluoroscopy table. Using AP fluoroscopy guidance, the T12 vertebral body should be identified. The fluoroscopic view should be rotated to a 45 degree oblique view to place the shadow of the body lateral to the laminae. A caudocephalad view may be necessary to correct for the lower thoracic kyphosis (approximately 20 degrees). The edge of the diaphragm lateral to the vertebral body is viewed. Its movement during inspiration and expiration are noted. If the diaphragm shadows the T12 vertebra and its rib, then the T11 rib is identified.48 A 10- to 15-cm blunt, curved cannula with a 10-mm active tip is introduced through a 16-gauge angiocatheter that has already pierced the skin. The target of the angiocatheter is at the junction of the rib and the vertebra.48 A tunnel view should demonstrate the cannula advancing toward the margin of the lamina. It is advanced along the lateral body until the tip is at the junction of the anterior one third and posterior two thirds of the lateral surface of the vertebral body.48 The patient should deeply inhale and exhale to demonstrate lack of penetration of the diaphragm in both the AP and lateral views. Proper needle placement must be assessed with the spread of the contrast medium (approximately 5 mL). This should demonstrate paravertebral spread under the pedicles in the AP view without intravascular, pleural, epidural, or intrathecal uptake. It should also flow medial to the interpleural space, above the crus of the diaphragm and anterior to the foramen.48 Sensory stimulation must be performed as described above (50 Hz and 0.9 V). The patient may report that he or she feels stimulation in the epigastric region. This is typical and satisfactory. If the stimulation is in a girdle-like fashion around the intercostal spaces, the needle needs to be pushed anteriorly.48 Motor stimulation must be performed and demonstrated to be negative as noted above. Additionally, 3 to 5 mL of a local anesthetic-steroid mix are injected after negative aspiration. A 1- to 2-minute delay should be observed before lesioning. Two lesions are performed with the active tip turned cephalad for the first lesion and caudad for the second. The lesions should be made at 80°C for 90 seconds. A postoperative end-expiratory chest radiography is taken to rule out a pneumothorax.

Raj et al48 reported on a group of 22 patients with RF lesioning of the splanchnic nerves. The technique was effective for all 10 patients with cancer in this group and complication free. Of the 12 patients in the nonmalignant pain group, some required repeat lesioning at 4 months. In this group, no complications were reported.48

Lumbar Sympathetic Chain

The lumbar sympathetic chain lies at the anterolateral border of the vertebral bodies. The location and size of the ganglia are variable.49 These ganglia consist of preganglionic sympathetic axons from T10 to L3 and post-ganglionic cell bodies. Classic teaching suggests the ganglia are located on the lower third of L2, upper third of L3, and mid-L4 and L5 vertebral bodies.21 Indications for lumbar sympathetic interruption are similar to the indications for stellate ganglion block, except the location of the pain is in the lower extremity:

• 1. Complex regional pain syndrome I (reflex sympathetic dystrophy)
• 2. Complex regional pain syndrome II (causalgia)
• 3. Herpes zoster
• 4. Postherpetic neuralgia
• 5. Phantom limb pain
• 6. Hyperhidrosis
• 7. Cancer pain to lower extremity
• 8. Acute radiation neuritis
• 9. Post-radiation neuralgia.

As with RF of the stellate ganglion, a diagnostic block must be performed prior to lesioning of the lumbar sympathetic chain. Complications that may occur with this procedure include:

• 1. Vascular injury
• 2. Intrathecal injury
• 3. Nerve root injury
• 4. Intradiscal placement of needle
• 5. Renal trauma
• 6. Punctured ureter
• 7. Genitofemoral/lumbar plexus neuralgia.

Aside from the complications, there are other concerns with lumbar sympathetic interruption. One concern is incomplete sympatholysis, which can make RF lesioning of the lumbar sympathetic chain less effective than phenol neurolysis.50 For more effective sympatholysis, at least three levels must be performed. Cllinical experience has shown a lesion at the L5 level may provide relief for foot pain.21 The other concern is the aberrant regeneration of the sympathetic nerve fibers over time. Future studies of this phenomenon would be helpful.

For the procedure, the patient is placed on the table in the prone position. The lumbosacral area is prepared and draped in a sterile manner. Using fluoroscopic guidance, in the AP view, the desired vertebral body is identified. The operator may need to adjust the fluoroscope in a craniocaudal fashion until the disc spaces are clear. The beam is then rotated in an oblique fashion to approximately 30 degrees (Fig. 74-4A). The desired view is one where the transverse process of the vertebra in question disappears behind the body.21 If the desired level is L2, 4, or 5, a skin wheal is raised at the inferior aspect of the transverse process at the lateral border of the vertebral body. If the desired level is L3, then the skin wheal is raised at the superior aspect of the transvese process at the lateral border of the vertebral body. A 16-gauge angiocatheter is directed through the skin wheal in “gun-barrel” fashion. A 15-cm, 10-mm active tip RF needle (a Racz-Finch curved, blunt needle from Radionics is suggested) is advanced through the angiocatheter and is advanced to the anterior border of the vertebral body, as noted in lateral view (see Fig. 74-4B). Care must be taken while advancing the needle so that it “hugs” the periosteum. Confirmation of needle placement may first be noted with the spread of 1 to 3 mL of water-soluble, nonionic contrast. The spread of this dye should be hugging the anterior aspect of the vertebral bodies in the lateral view. Proper placement for the particular level is also observed in the lateral view. At the L2 level, when the needle approaches the anterior border of the vertebral body, it should be at the junction of the superior two thirds, inferior one third of the body in the craniocaudal orientation. At the L3 level, the needle should be at the superior one third and inferior two thirds. At the levels of L4 and L5, the needle should be at the junction of the superior 50% and inferior 50%. In the AP view, the spread should appear vacuolated and remain under the pedicles. There should be no observance of myoneural spread as this indicates the needle remains in the psoas major muscle. (see Fig. 74-4C) Further confirmation of proper needle placement is accomplished with sensory and motor stimulation as described above. After assurance of proper placement of the needle, 1 to 3 mL of local anesthetic/steroid (at Texas Tech, this consists of a 50:50 mix of 2% lidocaine with 0.2% ropivacaine, and 1 mL of 40 mg/mL triamcinolone) should be injected. A 45-second to 1-minute delay should follow the injection prior to lesioning at 80°C for 90 seconds. The first lesion is with the active tip turned cephalad and the second lesion is with the tip turned caudad.21

Radiofrequency denervation of the lumbar sympathetic chain can be an effective treatment for persistent sympathetic-mediated pain.21 Lesioning the sympathetic chain at a single level (L3) only produces a 25% lasting response.21,51 Performing the procedure at multiple levels results in a 75% response at 8 weeks.21,51

Lumbar Facets

Zygapophyseal joints have been increasingly recognized as a potentially significant source of low back pain.53–55 The L1-4 dorsal rami form three branches—medial, lateral, and intermediate—in the intertransverse space.56 The L1-4 medial branches curve around the root of the superior articular process and pass through a notch bridged by the mamillo-accessory ligament. This branch supplies the lower pole at the same level and the upper pole of the joint above. It also supplies the rami in the multifidus.56–58 Each joint receives its nerve supply from the corresponding medial branch above and below the joint.56–58 Patients who have low back pain associated with localized tenderness right over the zygapophyseal joints without other root tension signs or neurologic signs, and who respond to intra-articular joint injections or to blocks of the medial branches of the dorsal rami, can be suspected of having zygapophyseal joint pain.59,60 Mooney and Robertson53 demonstrated referred pain from the zygapophyseal joints to the low back region, the greater trochanter, the posterolateral thigh, the groin region, and in a few patients, pain extended down the length of the leg and foot.53,60 These pain referral patterns were confirmed by Fukui et al.60 Complications of denervation of the lumbar facet include:

• 1. Increased pain
• 2. Neuritis
• 3. Numbness.

For the procedure, the patient is placed in the prone position. The lumbosacral area is prepared and draped in a sterile manner. The appropriate level is identified by fluoroscopic guidance in the AP view. The camera is rotated to a 10- to 15-degree oblique view, which allows visualization of the “Scottie dog” (Fig. 74-5A). A skin wheal is raised over the “eye” of the Scottie dog. A 16-gauge angiocatheter is advanced through the skin wheal to periosteum at the inferior aspect of the superior pars (the ear of the Scottie dog). A 10-cm, 10-mm active-tip RF needle is advanced through the angiocatheter to periosteum, then walked off in the lateral superior direction. In the lateral view this should be noted to be at the level of the facet line, posterior to the foraminal line, and below the level of the disk (Fig. 74-5B). Accurate placement is determined through sensory and motor stimulation: for sensory stimulation at 50 Hz and 0.9 V; for motor stimulation at 2 Hz and 2 V. Motor stimulation at 2 Hz will ideally cause paraspinal contractions, and at 2.0 V will not cause lower-extremity fasciculations.21 Lesioning is preceded by injection of 1.5 mL of a local anesthetic-steroid mix. After a 45-second to 1-minute delay, lesioning commences at 80°C for 90 seconds, medially and laterally, with an impedance, at less than 250 Ω.

The technique has had varying results.53,54,61–66 Most recently, however, Van Kleef et al67 performed a study of 31 patients with a 1-year history of chronic low back pain with positive diagnostic blocks. The patients were divided into a RF group (n = 15) and a control group (n = 16). The RF group received an 80 degree RF lesion; the control group had the same procedure but no current. The treating physician and the patients were blinded to group assignment. At 8 weeks of treatment, 10 of 15 patients in the RF group were deemed successful. There were 6 in the control group that also had pain relief at 8 weeks. However, at 3, 6, and 12 months, there were significantly more successful patients in the radiofrequency group compared with the control group.67

Cervical Facets

There are several possible pain generators in the cervical spine—the disk, the root, the zygapophyseal joint, or the muscles. These generators may result in pain that is experienced as a headache or as neck and arm pain. Many times, there can be poor correlation between the results from CT, magnetic resonance imaging (MRI), or radiologic studies and the pain suffered by the patient.21 A thorough history and physical examination can be helpful in the identification of pain related to the cervical zygapophysial joints. Common pain-referral patterns were established with the intra-articular injection of contrast medium.68 Cervical disc pain often goes along with facet pain, either as a result of degeneration or as a result of trauma. This may be the result of a sensitization of the annulus fibrosis nerve roots.21 In a study of 56 patients, 41% had both disc and facet pain, 23% had facet only, and 20% disc only.69,21 Diagnostic facet blocks offer the most success in selecting radiofrequency candidates, with a 70% response rate.70,21 Increasingly, the mechanism of chronic pain after whiplash injury has been deemed cervical zygapophyseal joint pain. This has been confirmed in biomechanical studies.71–75 Postmortem studies have revealed that the joints can be affected by pillar fractures, subchondral fractures, tears of their meniscoids, and intra-articular hemorrhages.71,75–78 It is evident that any of these findings could be associated with long-term, chronic neck pain.

To denervate the cervical facet, the anatomy must be appreciated. The cervical zygapophysial joints are innervated by articular branches derived from the medial branches of the cervical dorsal rami.70,79 The medial branches curve medially, hugging the waists of their ipsisegmental articular pillars.70 Articular branches arise from the medial branch as it approaches the posterior aspect of the articular pillar.70,79 An ascending branch innervates the zygapophyseal joint above and a descending branch innervates the joint below. Consequently, each typical cervical zygapophysial joint receives dual innervation, from the medial branch above and from the medial branch below its location.70 Owing to the anatomical differences, it is important to note that these anatomical characteristics for the techniques described hold true for C2-3 and below.

There are two different techniques described. The first technique is with the patient in the supine position and described by Wenger.21 The level desired is identified under AP fluoroscopic guidance. The camera is then rotated to a 30-degree oblique view so the vertebral foramen may be easily visualized. The camera axis is then oriented 10 degrees caudo-cranially.21 The posterior border of the facets is palpated and marked on the skin to assist with needle orientation. The cranio-caudad level of approach is identified by holding a clamp over the lateral neck and the cannula is inserted in this plane at the premarked posterior border of the facet. The cannula is advanced with a slight anterior angle and the bone is typically encountered in 1.5 to 2.5 cm. Cannula progress is followed in the 30-degree oblique view, taking care not to advance in the direction of the foramen. Once on bone, an AP view is taken to confirm placement at the waist of the body. Sensory stimulation is monitored for stimulation in the region of the neck and the shoulder. Motor stimulation should reveal the absence of upper-extremity motor fasciculations. An injection of 1.5 mL of a local anesthetic-steroid mix is given after negative aspiration. A delay of 45 seconds to 1 minute should occur prior to lesioning. The lesioning is then performed at 80°C for 90 seconds. Possible complications include neuritis and segmental nerve injury.21 This resembles the technique described by Sluijter and Koetsvelt-Baart.80

A second technique is with the patient in the prone position and is described by Lord et al.70 The level to be lesioned is identified in the AP view. The targeted area is the lateral margin of the articular pillar. A skin wheal is raised over the target and a 16-gauge angiocatheter is placed in gun-barrel fashion through the angiocatheter. The RF needle needs to be 10 cm with a 5-mm active tip. A needle with a curved, blunt tip is suggested for safety and ease of placement. The RF needle is advanced just medial to the lateral margin of the articular pillar until bony resistance of the back of the pillar is encountered.70 It is imperative that bony contact be made to prevent overinsertion. After the back of the pillar is contacted, the needle is walked off laterally in small increments, until a loss of bony resistance is sensed and the needle gently slips forward, tangential to the pillar.70 The patient is tested for motor and sensory stimulation as described earlier. Similarly, an injection of local anesthetic with steroid is given with a delay. The lesion is made at 80°C for 90 seconds, with the bevel curved cranially, then caudally.

In a study by Lord et al81 the efficacy of denervating the cervical zygapophyeal joints was examined. In this double-blind, placebo-controlled study, patients treated with RF had a median time of 263 days of at least 50% pain relief. Of the patients receiving RF, 41% reported numbness in the area of the treated nerves, but this was not deemed a problem.81

Lumbar Dorsal Root Ganglion

It has been demonstrated that compression of the dorsal root ganglion by a herniated disc generates radicular pain leading to prolonged repetitive firing in sensory axons. This leads to increased mechanical sensitivity.82 It has also been demonstrated that injury of nerve roots by chronic compression of the herniated disc, by previous surgery or by scar tissue itself, can lead to repetitive firing. Despite being compressed, stretching of these injured nerve roots leads to repetitive firing.83 Performing a ganglionectomy can be a decision based on diagnostic nerve root blocks that suggest a monoradicular pain syndrome.84 Ganglionectomy may be performed percutaneously by radiofrequency. The dorsal root ganglion is easily accessible by percutaneous needle. It is anatomically distinct from motor fibers and may be tested for placement with motor and sensory stimulation. A troublesome side effect of RF is neuritis. This may resolve by itself over a period of 4 to 6 weeks, or it may be treated with local anesthetic and steroid root sleeve injection. More recently, a pulsed radiofrequency technique has been advocated. While making radiofrequency lesions, the tip of the electrode is exposed to an electromagnetic field (EMF). The EMF may be the reason for the clinical effect of RF thermocoagulation. Unlike RF, EMF is not neurodestructive. Neuroablation occurs at greater than 45° C.42 EMF is performed at 42° C. A pilot study was done to investigate the usefulness of future double-blind studies with pulsed radiofrequency adjacent to the dorsal root ganglion in patients with unilateral symptomatology. Although the mechanism remains unclear, the study demonstrated improvement in the patients, and the conclusion was that EMF may be the clinical effect. Advantages of pulsed RF over RF: (1) a virtually painless procedure, (2) no signs of neurodestruction, and (3) may be suitable for patients with neuropathic pain or peripheral neuritis.85

To perform the procedure, the patient is placed in the prone position. The lumbosacral area is prepared and draped in a sterile manner. Using fluoroscopy, the appropriate level is identified in the AP view. If the L2 dorsal root ganglion is the desired level, then the superior pars of the L3 vertebral body is the target site. The camera is then rotated approximately 20 to 25 degrees oblique to the ipsilateral side. The camera is then given approximately 20 degrees caudo-cranial rotation. This assists with lining the discs. A skin wheal is made over the target area, which is the superior aspect of the superior pars. A 16-gauge angiocatheter is passed in gun-barrel fashion through the angiocatheter. A 10-cm, 10-mm, active-tip RF needle is passed through the angiocatheter. (It is suggested for technical and safety reasons that the RF needle be a curved blunt tip.) It is imperative the RF needle touch periosteum at the superior tip of the superior pars. The bevel of the needle is then rotated cranially and laterally, then gently walked off the superior pars. The bevel of the needle is then immediately rotated medially. The depth is checked in the lateral view at the 6 o’clock position in the posterior aspect of the foramina and at the level of the disk. In the AP view, the tip should be no more medial than the midfacet line.21 Sensory stimulation should be performed at 50 Hz. The patient should note a paresthesia in the distribution of the desired root between 0.3 and 0.7 V (preferentially, 0.3 to 0.5 V). If the stimulation can be felt at less than 0.3 V, there is concern for intraneural placement. If the stimulation requires more than 0.7 V, then the needle may be too far from the dorsal root ganglion. Motor stimulation should be negative at two times the sensory threshold to confirm a safe distance from the anterior motor fibers.21 Further confirmation of placement may be done with the injection of 2 to 3 mL of water-soluble, nonionic contrast dye. This should show a distribution along the nerve root and into the epidural space. After being assured of placement of the needle, the pulsed radiofrequency may be done at 42°C for 120 seconds, a total of three cycles. No local anesthetic or steroid needs to be placed prior to the EMF. Pulsed radiofrequency offers an alternative treatment technique for patients with neuropathic pain where RF is usually a contraindication.86 This technique needs further investigation to determine efficacy; however, case reports appear to be positive.87

Cervical Dorsal Root Ganglion

Pain in the neck may be related to a discogenic problem. If this is the case, pulsed RF of a cervical dorsal root ganglion may be indicated. As in all procedures, before therapy a diagnostic block with local anesthetic and steroid must be performed. The following description is for cervical roots C3 and below. The patient is placed in the supine position. The neck area is prepared and draped in a sterile manner. After identifying the appropriate level in the AP view, the camera is rotated to a 30-degree oblique view to observe the foramina. The technique is the same as for the cervical facets in the supine position; however, the 5-cm, 5-mm, active-tip RF needle is advanced to the six o’clock position into the posteriormost aspect of the vertebral canal. This dorsal-most position is used to avoid damage to the anterior vertebral artery. Frequent AP views are obtained to assure the needle does not pass medial to the midfacet line.21 Sensory stimulation should be performed at 50 Hz and a paresthesia in the distribution of the nerve root should be illicited between 0.3 and 0.7 V. Again, the ideal stimulation would occur between 0.3 and 0.5 V. Motor stimulation is performed at twice the threshold for sensory stimulation and should be found negative. Confirmation may be done with the injection of 0.5 mL of water-soluble, nonionic contrast dye, which should be observed following the nerve root and into the epidural space. Pulsed radiofrequency is then performed at 42°C for 120 seconds for three cycles. As with pulsed radiofrequency of the lumbar dorsal root, more studies are indicated. This technique, however, appears hopeful and avoids the side effects of hypesthesia, hyposensitivity, or neuritis.

Radiofrequency provides a discrete, safe, reliable, and reproducible technique for long-term pain relief. For best outcomes, proper patient selection is critical. It is imperative that prior to the technique, a proper diagnostic block is performed. To maintain safety and reliability, the procedure must be performed under fluoroscopic guidance with proper sensory and motor stimulation prior to the lesion. In the hands of a properly trained physician, radiofrequency provides a safe and effective treatment for difficult to treat pain conditions.