Subscription is FREE for qualified healthcare professionals in the US.
10 Articles in Volume 9, Issue #5
Dextrose Prolotherapy for Recurring Headache and Migraine Pain
Diagnosis of Low Back Pain
Ethics, Education, and Policy: Relationship and Mutual Reliance
Human Chorionic Gonadotropin in Pain Treatment
Musculoskeletal Ultrasound
Painful Herpetic Reactivation and Degenerative Musculoskeletal Injury
Post-stroke Pain
Preventive Medications for Chronic Daily Headache
The Pathophysiology of Neuropathic Pain
Use of Pulsed Radiofrequency in Clinical Practice

Use of Pulsed Radiofrequency in Clinical Practice

A retrospective study showed that PRF is effective as a replacement for both epidural steroid injections and surgery in the treatment of radicular pain as well as cases of peripheral nerve damage.

Radiofrequency current was introduced as a modality to make therapeutic lesions at various target nerves throughout the body. Its mode of action has been presumed to be entirely due to its effect of producing thermal lesions in neural tissue. However, this mode of action has recently been challenged.1-22 This article summarizes the current state of development of pulsed RF lesioning in clinical practice. It describes in detail a technique for performance of a pulsed radiofrequency lesion of the dorsal root ganglion for treatment of radicular pain. Finally, it summarizes the current literature regarding efficacy.

The widespread use of RF current for treatment of spinal pain began in 1980 when Sluijter and Metha introduced a 22g cannula through which a thermocouple probe could be inserted.21 This allowed the procedure to be performed percutaneously with minimal discomfort. The mode of action of radiofrequency current was initially attributed to the thermocoagulation of nerve fibers. This was brought into question when Sluijter performed the first pulsed RF lesion in February of 1996. He suggested that the electromagnetic field rather than temperature was responsible for the analgesic effect.1 It was subsequently determined that the electric field in particular was more likely to be responsible for the effect.2 Since thermal destruction of nervous tissue takes place at 45°C, Sluijter was careful to choose lesioning parameters that would not produce a temperature greater than 42°C (38° is more typical). This meant that the dorsal root ganglion might be treated without risk of deafferentation pain.1,6 Traditional thermal lesioning of the dorsal root ganglion has also been associated with neuroma formation, allodynia and dysesthesias.3 Given the potential for nerve damage with heating of the dorsal root ganglion, pulsed RF (PRF) introduced a method of treatment for radicular pain that had potential for therapeutic efficacy without the attendant risks.

Pulsed RF is delivered in short (20msec) bursts twice per second followed by a quiet phase (lasting 480msec) in which no current is applied. This allows for heat dissipation thus keeping the tissue temperature below the neurodestructive threshold of 45°C. Pulsing the current also allows the power output of the generator to be substantially increased. The usual output of voltage in the continuous mode is 15-25 volts while a pulsed radiofrequency lesion is usually performed at 45 volts.21

The role of pulsed RF in clinical practice has been an issue of debate. The argument against its use has been that it is unnecessary due to the availability of continuous RF which already has scientifically validated studies showing its efficacy. This argument, while true for treatment of medial branches, does not hold up well when considering use of RF current for the treatment of radiculopathies and painful peripheral neuropathies. For both of these chronic and painful conditions, there is currently little to offer these patients. When one considers the benign nature of this treatment and its possibility of real relief, there is little reason not to offer it as a therapeutic option.

Pulsed RF can be used without concern of producing sensory or motor loss.1,3,6,7,21 Since there is no satisfactory surgical solution for patients with chronic radicular pain or peripheral neuropathy, this mode of treatment offers a potentially valuable alternative to maintaining patients on chronic opiate therapy. In addition, it may offer an advantage over epidural steroids when one considers the reports of devastating outcomes associated with their use.23-26

Figure 1a. Oblique view. Note the needle position just below the pedicle and about 1/3 of the way down the cephalo-caudal extent of the foramen. Also note the position of the SAP about 1/3 of the way across the vertebral body. Figure 1b. AP view. Note the location of the needle tip at approximately the six o’clock position relative to the pedicle. Also note its lower than normal location about 1/3 of the way down the foramen. Figure 1c. Lateral view. Note the location of the needle tip about 1/3 of the way across the foramen in the cephalo-caudad direction and about 1/3 of the way across the foramen in the dorsal/ventral direction.

Pulsed RF was originally thought to be a totally non-destructive procedure. However, experimental work shows this is not the case.2 Convention holds that the word lesion should not be used when referring to a pulsed RF procedure. However, Cosman and Cosman measured heat bursts in the neurodestructive range occurring in a thin layer of tissue immediately surrounding the electrode.2 This evidence would dispute the currently held belief that pulsed RF does not cause a lesion. The dictionary defines the word lesion as “a localized pathological change in a bodily organ or tissue,” and Sluijter feels that a pulsed RF procedure clearly meets this criteria.27 However, neither he nor others who have looked at this issue feel that this small layer of cellular destruction fully explains the clinical effect produced by a pulsed RF procedure.2,18,27

The use of the pulsed mode in clinical practice has been slow in gaining widespread acceptance of this modality. This may be due to the paucity of evidence showing clear therapeutic advantage over placebo during the early years of its use. However, over the past five years, there have been several studies that demonstrate clear advantage over placebo. The elimination of the CPT code for the use of pulsed RF in 2005 has created yet another obstacle. However, there are still ethical and legal means of billing for the procedure and will be discussed later in the article.

Mechanism of Action

Initially, the effect of radiofrequency current was thought to be due solely to thermocoagulation of the fibers transmitting nociceptive stimuli to the spinal cord.6,18,22 Conventional heat radiofrequency causes non-selective thermal damage to the offending nerve by creating an electrical field between the small, uninsulated electrode needle tip (connected to a voltage generator) and a large inactive dispersion electrode. The current flow induces movement of ions within the tissue that alternate at the same frequency as the radiofrequency current.6,7,18,28 The result of this current flow is friction which produces heat in the tissues surrounding the electrode tip. The resultant lesion is spheroid in shape with the long axis parallel to the needle tip. The size of the lesion is fairly predictable and is based on tip temperature, lesion duration, needle diameter, length of the active tip, tissue vascularity, and heat conductivity of the surrounding tissue.6,7,28 This type of lesion is placed between the nociceptive focus and the spinal cord.13,16

The concept that tissue destruction was the means by which RF current produced its effect was re-evaluated in light of certain findings that were inconsistent with this theory.1-22 For example, it was known that heating of the dorsal root ganglion for the treatment of acute radicular pain due to a disc protrusion produced only a short term stunning effect on sensory fibers, while the pain relief seemed to last much longer. In addition, the application of heat to the dorsal root ganglion was distal (peripheral) to the nociceptive input. Thus, production of pain relief was not dependent on production of a lesion between the source of nociception and the spinal cord. These observations provided supporting evidence that led to the development of the pulsed radiofrequency procedure.

If neuronal destruction induced by heat is not responsible for the clinical effect, what is? Unfortunately, no single theory has been offered that fully accounts for the observed effects. Currently, despite evidence of a mild ablative effect,2,29,30 it is believed that the electric field is responsible for the clinical effect. Rather than producing local effects surrounding the electrode as with continuous RF, pulsed RF seems to produce its effect proximal to where is energy is applied. Indeed, changes within the central nervous system have been observed in response to pulsed RF energy.

When pulsed radiofrequency energy is applied to the dorsal root ganglion, it induces changes in gene expression within the dorsal horn of the spinal cord. The rapidly alternating current produces an affect to alter pain transmission via activation of a protein called C-Fos. On a cellular level, animal studies have shown that exposure of the dorsal root ganglion (DRG) to pulsed RF current causes both early and late induction of the protein C-Fos in layers 1 and 2 of the dorsal horn bilaterally.8,11 These effects are not temperature dependent8,9,11 and seem to occur as a result of current fluctuations, not because of tissue heating.1,2,9 Other proteins are also produced in response to pulsed RF current.17,18 However, it remains unclear whether any of these changes are responsible for the observed therapeutic effect.18 In addition, it is believed that strong electrical fields alter the nerve cell membranes so as to affect nerve transmission. This theory is supported by evidence showing that pulsed RF induces changes in synaptic transmission and causes electroporation.2,18

The Dorsal Root Ganglion Procedure

Radicular pain is produced when the dorsal root ganglion (DRG) is compressed or chemically irritated.31 Chemical irritation can be caused by disc protrusion, but this is not a requirement. It is believed that when the nerve is exposed to material leaking from the nucleus pulposus, a chemical inflammation is produced resulting in sustained discharge from the nerve. While most patients will exhibit some form of nerve compromise on imaging studies, there is a subset of patients that do not. These patients will complain of radicular symptoms yet will have a “normal” MRI. In this group of patients, it is believed that the radicular pain arises from chemical irritation of the nerve root as a result of inflammogens leaking from a damaged disc.32

Conservative treatment options for radicular pain or radiculopathy include rest, physical therapy, other physical modalities (massage, acupuncture, TENS, heat, and cold) and medications such as anti-epileptic medications, tri-cyclic anti-depressants, anti-arrhythmics and opiates. If these fail, the traditional approach has been to perform a series of epidural steroid injections at the involved level and, if necessary, proceed to surgery. Surgery has traditionally been considered the backup treatment for radiculopathy.31 The rationale in performing surgery is that removal of the lesion causing the pain should relieve the symptoms. However, both epidural steroid injections and surgery are not without attendant risks.

Epidural injections have been advocated as a means of obviating the need for surgery. Long-term results have been reported—but not substantiated—by studies meeting the highest level of evidence.33-35 Reports of serious complications in the literature have been widely promulgated throughout the pain management community. There have been reports of spinal cord injury and death after transforaminal epidural steroid injections in both the cervical and lumbosacral areas. Some patients have been left with persistent paraplegia and quadriplegia as a result of these procedures.23-25 The theorized cause of these incidents is particulate steroid injected into an end artery that serves to support perfusion of the anterior spinal or vertebral arteries.24 There have also been reports of temporary loss of vision with incomplete recovery. This was thought to be due to either retinal or vitreal hemorrhage.26 Aspiration to identify when the needle has been inadvertently placed into a vessel is not a reliable test.36,37 When the local anesthetic and steroid are injected in separate syringes with the operator waiting 90 seconds between the injections, serious permanent complications have been avoided.36

Anatomy: There are five paired nerves that exit their respective intervertebral foramina from L1-2 to the L5-S1 levels. Just as the orientation of the lumbar zygapophyseal joint differs from L1-2 to L5-S1, the lumbar nerves exit their respective foramina at different angles from L1 through L5. At L1, the nerves exit downwards and forward at an acute angle whereas, at L5, the nerves exit more horizontally and at a more obtuse angle.6,10,38 The lumbar ventral roots find their cell bodies of origin within the spinal cord at the T 9-11 vertebral level.39 Rootlets come off the dorsal and ventral surface of the spinal cord to form the dorsal and ventral roots. The dorsal root ganglion contains cell bodies that provide sensation, proprioception and pain. The dorsal and ventral roots then join to form the spinal nerve root.10

The location of the DRG varies. In the lumbar spine, it is located from the mid to the anterior aspect of the foramen in a saggital plane 98% of the time and is extra-foraminal in 2% of cases. At the L4 level and above, the DRG is usually located within the foramen but, at L5, 5.7% of DRGs studied were intraspinal and 77.3% were intraspinal at S1.3,40 In the frontal plane, the DRG is located at the junction of the upper one third and the lower two thirds of the foramen.

The anatomy of the arterial system is important when performing spinal injections because some of the vessels that supply the spinal nerve roots also reinforce the blood flow to the anterior spinal artery. A branch of each vertebral artery combines to form the anterior spinal artery. At several levels throughout the spine, medullary arteries form to reinforce the blood flow to the anterior spinal artery.10 On the left side of the spine, between T8-L2, the largest medullary artery—termed the artery of Adamkiewicz—is found. There have been reports of central nervous system sequelae after performance of cervical, lumbar, or sacral nerve root blocks.23-25,36 In one report, three cases of paraplegia were thought to be due to needle perforation of the artery of Adamkiewicz that originated at an unusually low location. All three cases occurred after injection of local anesthetic and steroid combined in the same syringe.25 The exact cause of this outcome was unknown. One theory is that particles in the steroid embolized within the territory supplied by the artery of Adamkiewicz, producing spinal cord infarction.25 It is for this reason that injection of contrast dye under live fluoroscopy is recommended. One should monitor for a flash of dye seen within the central canal, indicating uptake of dye by the artery. A second physiological test is to inject local anesthetic first, wait 90-120 seconds, then make certain the patient can move all extremities prior to injection of the steroid.36

Figure 2a. The procedure can be done at any spinal level. Here is an open mouth view of a needle placed at C2 for treatment of occipital headache. The patient had previously responded positively on two occasions to occipital nerve blocks. Figure 2b. Here is the initial view showing correct needle placement for a C2 PRF. The needle is pointed directly at the targeted nerve. Figure 3a. Starting view of L5 DRG. Figure 3b. Lateral view prior to injection of dye. Figure 3c. Oblique view after dye injection.

Absolute contraindications3,28:

  • Patients who are unwilling or unable to give consent for the procedure.
  • Local or systemic infection: this could put the patient at risk for spread of infection to the central nervous system.
  • History of anaphylactic reaction to contrast dye: reports of spinal cord injury after injection of local anesthetic and steroid mixed in the same syringe gave rise to concern that some component of the injection may have clogged a radicular artery, reinforcing blood flow to the anterior spinal artery. Use of contrast dye injected under live fluoroscopic x-ray is thought to be essential to monitor for the spread of dye into the central canal.
  • Bleeding disorders due to disease or concurrently used medications: medications that may cause bleeding must be stopped at the appropriate time prior to the procedure.
  • Uncooperative patient: although this may not be discovered until attempting the procedure, it is worth discontinuing the procedure rather than causing permanent nerve damage.
  • Needle phobia that is not treatable by psychological interventions.
  • Relative Contraindications3,28:
  • Allergy to any medication planned to be used during the procedure.
  • Pregnancy.
  • Anatomical derangements that could prevent successful completion of the procedure: patients with previous spinal fusion should have an x ray to evaluate access.

Procedure Performance-DRG RF

Although this procedure can be done at all spinal levels, this discussion will be limited to the lumbar dorsal root ganglion. The approach is well described in the ISIS guidelines in the chapter on lumbar spinal nerve block. It is referred to as a retroneural approach.28 In this case, the target lies at the intersection of two lines. When viewed laterally, the first line runs longitudinally between the posterior and anterior half of the foramen, bisecting the foramen into two equal halves. The second line runs in a transverse direction between the superior one third and the inferior two thirds of the foramen. The intersection serves as a starting point for locating the dorsal root ganglion, however, it can lie anywhere between the mid aspect to the most anterior aspect of the foramen in the anterior-posterior plane.

To achieve the target point, an oblique view identical to that used for a transforaminal procedure is obtained (see Figure 1a). After squaring the superior endplate at the involved level, the image is obliqued until the superior articular process is projected one third of the distance across the image of the vertebral body (approximately 15-20 degrees). In this view, the starting point for the needle is slightly inferior and lateral to that used for a transforaminal injection. Aim for a point just underneath the pedicle, one third of the way down the foramen. As the needle is advanced, rotate into an AP projection to assess depth of insertion (see Figure 1b). If further insertion is required, rotate back to an oblique view and advance. When the needle tip approaches the lateral aspect of the vertebral body (see Figure 1c), it is best to advance further in an AP view. The needle should be advanced very slowly (only a millimeter at a time). This can be achieved by pinching the needle shaft at the point of skin entry. Patients should be warned that they will feel a paresthesia and not to jump. Because the greatest current density is projected from the tip of the needle, it is imperative that the needle be pointed directly toward nerve tissue and not against the bone of the vertebral body—i.e. the needle should be perpendicular to the target as opposed to parallel when the needle touches vertebral body. Once the paresthesia is felt, place the electrode into the needle and begin testing. The operator must be very careful when handling the needle at this time as any movement risks spearing the nerve.

Electrical stimulation tests are used to determine proximity between the needle tip and the nerve. Adequate placement requires that the patient feel reproducible stimulation (tingling of the stimulated dermatome in the lower extremity) at less than 0.25 volts but greater than 0.1V. Stimulation is done twice: once to determine the minimum sensory threshold (the lowest voltage at which the patient can still perceive a sensation) and a second time to determine reproducibility. When stimulating the second time, the voltage is slowly increased until the patient reports that they perceive a stimulus. This should be within 0.05 volts of the first stimulation test. If the patient does not feel the current at the required level of less than 0.25 volts, the physician should advance the needle slightly (no more than 1 mm) or reposition the needle altogether and re-test. Stimulation of the nerve must always be at a voltage of greater that 0.1V; anything less risks intra-neural placement.

If the operator is having difficulty locating the nerve, the needle tip is usually too medial and should be corrected in an oblique view so that the tip is located directly beneath the pedicle on a line that bisects the pedicle. It may also be misplaced too cephalad and should be redirected slightly inferior. There is no need for motor stimulation when performing pulsed radiofrequency since the modality does not damage motor fibers.

The next step is to check impedance. The maximum impedance should be less than 400 ohms and ideally less than 250 ohms. To achieve this, inject a small amount (1 ml) of saline or local anesthetic (2% xylocaine). Prior to injection, the needle must be retracted 1-2mm in order to avoid intraneural injection which can have devastating consequences, including irreversible sensory and motor loss. If there is any resistance during injection, stop, back the needle up slightly, and inject again; then replace the needle to its original position. Liquid should flow easily through a 20g needle. The initial minimal stimulation threshold (MST) and impedance (I) are recorded prior to treatment.

At this point, turn on the power in the pulsed RF mode. Slowly increase the voltage until the patient can verify that he/she feels the pulsing. The needle must be close enough to the target nerve tissue to produce a perceptible electrical discharge in the treated extremity with each pulse. In other words, during the time the pulsed RF mode is operative, the patient must report a sensation of pulsing in the appropriate dermatome. If the patient does not, it may indicate that the needle is not close enough to the targeted nerve tissue to produce an effect. If no pulsing is felt, the needle needs to be repositioned prior to treatment. Although no study has demonstrated this step to be necessary, it can be useful as another test to verify proximity to the targeted neural tissue. A routine procedure does not require anesthetizing the nerve. However, if performing the lesion causes pain, the nerve should be anesthetized prior to performing the procedure. This step should be avoided, if possible, for two reasons. First, it can result in intraneural injection as mentioned above and, second, movement of the needle risks losing correct position obtained by sensory stimulation.

Proceed with pulsed radiofrequency treatment for three to four minutes at 200 milliamps or 45 volts (as long as the temperature does not exceed 42°C), at two pulses per second, with current applied for 20 milliseconds each pulse. Treatment protocols may vary with different operators, and the author recommends consulting the literature for other examples of lesion parameters (see Figures 2a and 2b).

Post-treatment Patient Advice

The patient usually feels better immediately after completion of the procedure due to the injection of local anesthetic onto the affected nerve. When the affect wears off, the patient may begin to feel sore. Advise that he/she may continue to feel sore for the first week, better the second week, and that the full effect will usually take three weeks to develop. During this time, it is not required that the patient restrict activities, except as needed due to pain. Deep tissue massage once a week for the first three weeks following the procedure may relieve soreness due to the procedure itself, as well as chronic trigger points which may have developed over the course of the disease.


Bleeding and infection are commonly listed but are rare complications. Hematoma may occur just under the skin or in the deeper muscle layers. Vaso-vagal reaction may occur during procedure performance. Mechanical nerve root damage can occur from needle trauma but is rare. However, if fluid is injected into the axon bundle, this can leave the patient with persistent motor and/or sensory deficits. Starting in too oblique a position can result in intra-thecal placement, resulting in dural puncture and/or injection of local anesthetic into the intra-thecal space. This could cause a post-dural puncture headache or lower extremity paralysis. Intra-thecal placement may also occur if the needle is advanced beyond the six o’clock position in a patient with long dural root sleeves. All complications are extremely rare and there have been no serious, long-term complications due to pulsed RF reported in the literature.


In a prospective placebo-controlled trial by Van Zundert and colleagues, 23 patients underwent either a pulsed RF procedure adjacent to the cervical DRG or a sham lesion. Eleven patients were in the PRF group and twelve were in the sham group. At the three-month follow up, 83% of patients in the PRF group reported at least 50% improvement in global perceived effect, compared with 33% in the sham group. 82% of patients reported at least a 20-point decrease in VAS vs. 25% in the sham group. PRF patients also showed a decrease in medication use. At the six-month follow up, 64% of PRF patients reported at least 50% improvement in global perceived effect, compared with 17% in the sham group. Again, the PRF patients decreased medication use. Overall, the researchers felt that PRF provided better relief than sham intervention at three and six months after the procedure.15

Teixeira and colleagues reported on the use of PRF of the DRG for the treatment of radicular pain and radiculopathy.41 They retrospectively studied thirteen consecutive patients with pain originating at the L3 to S1 spinal levels. Of the patients, twelve were professionals who had stopped working due to the pain. The average numeric pain score on presentation was 7.82. All patients had failed various conservative treatments. Nine of the thirteen patients had slight to moderate loss of motor function, nine patients had sensory loss, and twelve patients had paresthesias. A diagnosis of herniated intervertebral disc, concordant with clinical presentation, had been made with MRI or CT scans in all patients. All patients had been seen by a neurosurgeon and were scheduled for surgery.

Prior to treatment, proximity to the targeted neural tissue was verified by sensory stimulation at 50 hertz. The mean stimulation threshold was 0.22 volts with a range of 0.08-.03 volts. Patients were treated with pulsed RF for three minutes at 45 volts, and two pulses of RF current for 20 msec each. At the final follow up examination, 15.8 months after the initial procedure, one patient had undergone disc surgery due to continued leg pain. The twelve remaining patients did not require surgery for leg pain, but one patient did undergo spinal fusion for back pain. Of the remaining patients, the average pain score was 0.27, and all twelve patients with professions had returned to work two weeks after the procedure. Neurological abnormalities resolved in eleven of the twelve patients. One patient had decreased sensation in a small area of the thigh innervated by the L3 dermatome. The average Numeric Rating Scale (NRS) score fell significantly one week after the procedure. The results were statistically significant at four weeks post procedure. Motor and sensory loss improved significantly two weeks post treatment.

Another retrospective study looked at the clinical effect of PRF lesioning of the DRG in patients with radicular pain of varying etiologies. There were three patient groups divided according to etiology of the radicular pain: Herniated Disc (HD), Spinal Stenosis (SS), and Failed Back Surgery Syndrome (FBSS). The results showed a decrease in numeric rating scale in patients with HD and SS but not FBSS. They concluded that PRF of the DRG was significantly more effective in patients with HD and SS than with FBSS. Of note: they do not mention their stimulation parameters prior to lesioning. They reported that 40% of patients in the HD and SS group reported “successful treatment” at 180 days post treatment.42

Ahadian reported on the use of PRF for a variety of painful disorders in a retrospective study. At three month follow-up, he found fair to excellent relief in 40% of patients with cervical radiculopathy, 64% of patients with lumbar radiculopathy, and 56% of patients with peripheral nerve lesions of varying sorts.12

Cahana and colleagues performed an extensive literature search on the topic of pulsed RF.5 They found 58 reports on the clinical use of the procedure noting mixed results. This is the most comprehensive review of the literature to date on the topic of pulsed RF and is recommended as a good review of the current literature (see Figure 3a-d).

Author’s Opinion

In reviewing the data presented above as well as from personal experience, this author concludes that the patients with stimulation parameters between 0.1-0.3 volts have the longest lasting success.

Values below 0.1 volts may reflect intraneural placement.30 Since there is evidence that a small lesion does occur around the electrode tip, it may be unwise to allow the electrode to penetrate neural tissue. For pulsed RF to be effective, the electrode must be carefully and meticulously positioned, pointing directly perpendicular and very close to the targeted nerve. The operator should have very soft hands and excellent needle handling skills before attempting this procedure. The patient must also be warned just prior to eliciting a paresthesia and told not to make any sudden movements. Before injecting any fluid, the needle must be retracted from the position intended for lesioning. Once the anesthetic has been injected, the original position can be re-established. When injecting fluid through the needle, its position must be secured by one hand at the skin and not be allowed to move, lest it cause a severe pain and possible needle trauma to the nerve. Also, no injection should be made if there is even slight resistance to flow of the fluid.

Billing Considerations

During the performance of the procedure, local anesthetic is injected to anesthetize the nerve prior to lesioning. The CPT code 64483 reads “Injection,….of diagnostic or therapeutic substance(s) (including anesthetic, …..).” Therefore, this code can be appropriately billed in this situation. Since there is no CPT code for dorsal root ganglion lesion, the physician may decide to add a small charge to cover the cost of supplies and other overhead. In other words, the transforaminal injection and the dorsal root ganglion injection are billed as two separate procedures. If one desires to add an extra charge paid by the patient, it would be wise to check with the directors of the various local insurance companies prior to billing for the dorsal root ganglion portion of the procedure in this manner.

Peripheral Nerve PRF Procedures

Peripheral nerve damage can occur in a number of ways, including previous surgery to the area, compression or crush injuries, and various chemical etiologies. Cohn and Griffith state, “following peripheral nerve injury, ion channel modulation occurs leading to nociceptor sensitization, expansion of receptive fields, diminished central inhibition, increased neuronal excitability in the spinal cord, and reorganization in the dorsal horn.”3 The results are hyper-excitability and increased spontaneous firing of damaged nerves. There are currently few options for patients suffering from peripheral nerve damage, and the available methods are complicated, risky, and/or expensive. PRF is a relatively simple, inexpensive procedure that can produce long-term pain relief without the risks that frequently accompany other methods.

Figure 3d. AP view after injection of dye. Figure 4a. This is an image of a 5cm needle directed at the suprascapular nerve located within the notch.

The general principles for treatment of peripheral nerve injury using PRF are similar to those described in the techniques for treatment of the DRG. As in the DRG procedure, the electrode must be meticulously positioned directly perpendicular to, and very close to, the targeted nerve. However, when performing peripheral nerve lesioning, the nerve is usually located using surface landmarks and by palpation of the artery associated with the nerve. In most cases, fluoroscopy is not used. This can make precise localization more difficult. Fortunately, the use of ultra-sound for localization of peripheral nerves is becoming increasingly more common. This holds the promise of making precise localization much easier.

The technique can be used on a number of peripheral nerves including the occipital nerve, lateral femoral cutaneous nerve, ilioinguinal nerve, tibial nerve, and others. The anatomy and indications vary depending on the involved nerve. These are well described in Waldman’s Atlas of Interventional Pain Management.43 These procedures can also be more problematic than the DRG procedure because most peripheral nerves are not very deep in the tissue. This results in the needle wobbling within the tissue and makes it difficult to hold the tip in the precise position over the nerve, especially because the head of the needle is heavy relative to the tip. It requires a great deal of attention and concentration to keep the tip pointed directly over the targeted nerve.

Typical Peripheral PRF Procedure: Suprascapular Nerve Treatment Using Radiofrequency Aided by Fluoroscopy


The suprascapular artery and vein travel with the nerve. The suprascapular nerve originates from the ventral ramus of C5 and C6. It travels inferiorly, posteriorly, and laterally to run under the transverse scapular ligament before entering the suprascapular notch. It is at this point that the nerve is accessible to block and PRF. The nerve then travels to the muscles of the shoulder girdle and the posterior joint capsule. The nerve is a mixed motor and sensory nerve that also contains nociceptive fibers. Seventy percent of the sensory information to the shoulder girdle is provided by this nerve.43,44


The procedure is indicated for patients with shoulder pain who have been evaluated by a shoulder specialist and found not to be a candidate for surgery. Many of these patients have had a prior rotator cuff repair and still have pain. It is also indicated for patients with a frozen shoulder from any cause. Patients must have responded positively (70-100% reduction in pain) to blockade of the nerve with local anesthetic on two separate occasions to be candidates for PRF.44


Fluoroscopy is used to perform the procedure. The patient is placed in a prone position on the table. It can sometimes be difficult to visualize the suprascapular notch. The technique is to angle the c-arm in a cephalo-caudal direction and then oblique towards the affected side. The suprascapular notch is identified just above the spine of the scapula and medial to the coricoid process.44 After sterile prep and drape, a 5 cm RFK needle with a 4 mm active tip is advanced down the angle of the x-ray beam in order to position the needle perpendicular to the nerve (see Figures 4a and 4b). Next, the nerve is stimulated at 50 hertz with the intention of obtaining a paresthesia in the shoulder at less than 0.2 volts. This should be repeated to test for reproducibility. It is not necessary to perform motor stimulation, as PRF does not affect motor fibers. Then the impedance is checked. It should register at less than 400 ohms. If not, saline is injected. It is important not to numb the targeted nerve in order to ensure that the patient feels continued pulsing throughout the entire procedure. Finally, the nerve is lesioned in the pulsed mode for four minutes at 2 hertz, and 45-60 volts. It is not necessary to measure temperature since the procedure, as described above, produces temperatures below 42°C. The operator should hold the needle in place throughout the procedure and occasionally ask the patient if they continue to feel pulsing. If not, it implies that the tip has moved away from the targeted nerve and is no longer producing any effect. A visible pulsation of the supraspinatus and infraspinatus muscles will be seen.44

At the completion of the procedure, ½-1 ml of 3% phenol can be injected through the needle if the operator plans to bill insurance for the procedure. This concentration is adequate to cause neurolysis of the small pain fibers without damaging the motor or sensory fibers.

Complications and Efficacy of Peripheral Nerve PRF

Side effects and complications are minor and vary according to the location of the procedure being performed. The efficacy of most procedures has only been studied retrospectively but with promising results. For example, the Ahadian study found fair to excellent relief in 56% of patients on whom various peripheral nerve PRF procedures were performed.12 Given that there is not much else to offer these patients (short of spinal cord or peripheral nerve stimulation), these results warrant attempting this simple and inexpensive procedure before deciding on more costly remedies.


Although the clinical evidence showing its efficacy is still accumulating, there is currently enough evidence to recommend use of PRF in the treatment of pain.29 At this point, its best potential is in the treatment of radicular pain and peripheral nerve damage. The previously reported prospective study showed promising results. In the author’s opinion, the results would likely have been improved if the needle had been placed in closer proximity to the targeted nerve. A retrospective study showed that PRF is effective as a replacement for both epidural steroid injections and surgery in the treatment of radicular pain. PRF can be used in cases of peripheral nerve damage as a safe alternative to more aggressive treatments such as chemical neurolysis, which can result in further injury. Additionally, PRF is considerably less expensive than spinal cord and peripheral nerve stimulators, which have traditionally been options. As new studies are conducted and the modality is more widely utilized in clinical practice, new uses will be uncovered.

Last updated on: January 5, 2012
close X