Treatment of pain, acute, subacute, or chronic, often involves a continuum of treatment options. Initially, conservative options such as physical therapy, medications, and alternative/complimentary treatments should be considered. If these modalities provide inadequate pain control, interventional procedures, which inherently carry more risk, can be considered. In this chapter, we review relevant anatomy, theory, indications, and provide an overview of how common pain procedures are performed. This chapter is not meant to be a comprehensive guide or instructional manual. The landscape of interventional pain practice continues to evolve and change; as such we highly recommend that providers pursue an accredited fellowship in interventional pain, and stay current on the latest techniques, limitations, and literature. Given this, we present a broad outline and theoretical basis of commonly utilized pain procedures.
Lumbar Medial Branch Blocks
Lumbar facet (zygapophyseal) joints (z-joints) are an important potential source of low back pain. The most frequent etiology of lumbar facet pain derives from degenerative disease. Surgical fusion of two vertebrae can also lead to accelerated degeneration of adjacent levels of facet joints. , Another frequent cause is trauma secondary to whiplash injuries, although more common in the cervical region. ,
Degenerative facet arthropathy is the most common cause of facet-mediated pain and occurrence increases with age. Up to 40% of elderly patients with chronic low back pain have facet-mediated pain. As such, diagnostic confidence of a medial branch block is increased when performed on an older patient with chronic low back pain in comparison with a younger patient, who may have other sources of low back pain . Axial, nonradicular low back pain may be consistent with pain from the facet joints. Facet pain referral patterns have been well studied, and although they can vary, typically follow predictable patterns. Pain radiating below the knee is a negative predictor of facet-mediated pain , . Stiffness, pain worsened with prolonged standing and relieved by recumbency is consistent with lumbar facet pain . Imaging, including X-ray, CT, and MRI, is helpful in excluding other diagnoses; however no radiographic finding is sensitive or specific for facet-mediated pain. The correlation of facet joint arthropathy and a positive response to diagnostic medial branch blocks is not clear. , Given the ambiguous nature of diagnosing facet-mediated pain via patient history, physical exam, and radiographic findings, medial branch blocks have become the gold standard for diagnosis.
Facet joints are paired synovial joints located posterolateral to the vertebral body and comprised of the superior articular process at and the inferior articular process below that level. The two facet joints and intervertebral disc at each level make up the “three-joint complex” which stabilizes and supports the spine. In the lumbar spine, the facet joints are generally oriented in the sagittal plane. The sensory innervation to each lumbar facet joint is supplied by the medial branch of the posterior rami. The medial branch of the posterior rami also innervates the multifidus muscle, interspinous muscle and ligament, and the periosteum of the neural arch. Each facet joint receives dual innervation from the medial branch at, and the medial branch above, the level of the facet joint. For example, the L4–L5 facet joint receives its innervation from the L3 and L4 medial branch nerves. Each of the L1–L4 medial branches travels across the transverse process at the junction of the transverse process and the superior articular process. As the nerve passes medially, the medial branch is covered by the mamillo-accessory ligament which can become calcified. Due to its course and the mamillo-accessory ligament, the target for the lumbar medial branches is the junction of the transverse and superior articular process ( Fig. 13.1 ). A variant of innervation of the facet joints in the lumbar spine is the innervation of the L5–S1 facet joint, which is innervated by the medial branch of L4 above and the L5 dorsal ramus. The L5 dorsal ramus travels along a course where it is accessible at the junction of the superior articular process of S1 and the sacral ala.
Theoretical Basis of Procedure
Due to the high prevalence of false-positive responses to a single medial branch block, two confirmatory blocks done on separate occasions should be performed. , Although there is some debate, all guidelines and position papers to date by major interventional pain and spine societies recommend a double block paradigm. The block should produce significant relief following each injection. Criteria for success per Spine Intervention Society Guidelines are 80% relief of pain. To avoid false-positive responses to blocks, low volumes of anesthetic are used, typically between 0.3 and 0.5 mL of volume. The concern with higher volumes is that anesthetic may spread to block adjacent nerves, including the lateral and intermediate branches which innervate the paraspinal muscles, ligaments, sacroiliac joints, and skin. False-negative responses due to vascular uptake, and the low sensitivity of aspiration prior to injection, argue for the use of contrast prior to injection of the anesthetic. The dual innervation of each level of facet joint requires the block of both medial branches innervating each joint (e.g., L3 and L4 medial branches for the L4–L5 facet joint) to achieve a successful block.
Procedure: L1–L4 Medial Branch Blocks
The patient is placed in the prone position and an anteroposterior (AP) view of the desired vertebral level is obtained in the center of the X-ray beam. The beam should be tilted cephalad or caudad to line up the superior endplate at the target level. Next, the C-arm is obliqued in the ipsilateral direction to optimize the junction of the superior articular and transverse processes. There should be clear visualization of the transition area between the transverse process and the superior articular process. The target for final needle placement should be the junction of the superior articular and transverse process ( Fig. 13.2 ). 22- or 25-gauge 3 ½-inch or 5-inch spinal needles are inserted into the skin with close attention to the desired trajectory to the optimal endpoint at the inflexion of the superior articular process and transverse process. 25-gauge needles may be used without skin anesthesia. If the target is well-identified and the needle insertion site is well-chosen, this can be efficiently achieved by inserting the spinal needle coaxially with the intensifier, i.e., the spinal needle should be parallel with the beam of the intensifier. Careful attention should be paid to prevent the needle from migrating cephalad to the transverse process, so as to not enter the intervertebral foramen. After aspiration, a small aliquot of contrast is injected under live fluoroscopy to confirm location and ensure there is no vascular uptake ( Fig. 13.3 ). If vascular flow is detected, the needle should be redirected and confirmed that contrast remains over the medial branch. Once confirmed, 0.3–0.5cc of anesthetic is injected at each level and the needles removed.
Procedure: L5 Dorsal Rami Block
For the L5 dorsal ramus, the target is the groove between the sacral ala and the superior articular process of the L5–S1 joint. The dorsal rami run caudad from this groove, ultimately splitting into the lateral and medial branches. Thus, this block is of the dorsal rami prior to the division. On fluoroscopic view, the target is generally best viewed with ipsilateral obliquity in order to move the superior articular process medially in relation to the sacrum, thus exposing the target. Particular attention should be paid to the iliac crest when optimizing this view as it may obscure direct access to the target. Once the target is visualized and identified, a 22- or 25-gauge spinal needle is inserted and advanced to the target using coaxial alignment. A small aliquot of contrast is injected to confirm location and ensure no vascular uptake. If vascular flow is noted, redirect the needle and reinject to confirm that contrast remains over the dorsal rami. Once confirmed, 0.3–0.5cc of anesthetic is injected and needles are removed.
Cervical Medial Branch Blocks
Cervical facet (zygapophyseal) joints (z-joints) are potential causes of neck pain, shoulder pain, upper back pain, and headaches. Degenerative disease and whiplash injuries account for the majority of facet-mediated pain in the cervical spine.
Neck pain referred from the facet joints follows predictable patterns that correspond to levels of facet joints. Each joint is innervated by the medial branches of the cervical dorsal rami. For those suffering with chronic neck pain, the most common cause is of facet joint origin and the prevalence of facet-mediated chronic neck pain has been reported as high as 55%. The most common cervical joints identified as the source of pain have been C2–C3, C5–C6, and C6–C7, respectively. Studies conducted by Bogduk, Dwyer, and Aprill originally mapped referral patterns of pain from the different facet joint levels. This distribution of pain has been confirmed in other studies and serves as the basis for targeting joint levels. These pain referral patterns are useful when the pain is thought to be derived from the facet joint; however, these referral patterns do not necessarily exclude other potential causes of neck pain. Headaches associated with upper cervical joint disease are indicative of the C2–C3 facet joint. , Although there is some debate, all guidelines and position papers to date by major interventional pain and spine societies recommend using a double block paradigm.
The innervation of the facet joints C3–C4 to C7–T1 receive dual innervation from the medial branches of the posterior rami at the same level and the level below (e.g., the C6–C7 facet joint is innervated by the C6 and C7 medial branch). The path of the medial branches innervating C3–4 to C7–T1 is targeted as they cross the waist of the articular pillars. In general, a diagonal line should be drawn from corner to corner of the articular pillars. The intersection of these lines indicates the target for the medial branch block. The medial branch of the C2 dorsal rami, also known as the greater occipital nerve, receives a communicating branch from the third occipital nerve (TON), and together with a separate articular branch of the third occipital nerve innervates the C2–C3 facet joint. This level and the C7 medial branch differ in their approach. The third occipital nerve is accessible to block at the junction of the C2 and C3 articular pillars, and generally will be adequately blocked with two targets: just above and just below the C2–3 joint line, taking care not the enter the joint itself. The C7 medial branch is much more variable in its path and often does not cross the lateral mass in the same fashion as the other cervical medial branches. Instead the medial branch of the C7 dorsal rami courses over the superior-lateral transverse process and can vary in its course.
Theoretical Basis of Procedure
Medial branch blocks are diagnostic in nature; they do not provide long-term pain relief and are a means to an end, which is radiofrequency ablation (RFA). If the medial branches, which supply innervation to a painful joint, are anesthetized and provide pain relief, one would expect that destruction of those nerves with thermal RFA to provide longer-term relief. One of the most important aspects of medial branch blocks is communication with patients and ensuring that patients understand the purpose and expectations of the procedure. These blocks should only be done if patients are open to an ablative procedure, which might provide longer-term relief; if not, pursuing diagnostic medial branch blocks is a futile effort.
The patient may be positioned in either prone or lateral decubitus, depending on operator preference. A true lateral is necessary for procedure safety and proper placement of needles. For the medial branches of the C3 through C6 vertebral levels, the target is the center of the lateral cervical mass, which can be visualized as the intersection of two lines each going diagonally from the superior corner of the articular pillar to the opposite inferior corner. The target for the third occipital nerve is at the junction of the articular pillars of C2 and C3. The C7 medial branch dorsal rami target is the superior articular process. Given the smaller target for C7, it is imperative to keep the needle over the target of the superior articular process as to not go too ventrally and enter the C7–T1 intervertebral foramen or too dorsally and enter the C6–C7 facet joint. AP views confirm that the needle tip has not moved medial to the lateral mass. The needle target should be placed in the center of the image to avoid imaging parallax. Once identified, a 25-gauge 2 ½ or 3 ½ inch spinal needle, depending on patient body habitus, is introduced into the skin and advanced in coaxial view to the target. A small amount of contrast (0.3 mL) is then injected to ensure lack of vascular uptake. After confirmation, 0.3–0.5 mL of local anesthetic is injected with careful stabilization of the needle against os and the needle is removed.
The target of RFA is also the medial branch, and as such anatomy is the same as outlined in the medial branch section above.
Theoretical Basis of Radiofrequency Ablation
RFA of the medial branches of the primary dorsal ramus has long been accepted as a safe and effective way to treat axial low back pain due to painful facet joints. The theory behind the procedure was initially described by Shealy in 1974 after reviewing the results presented by Rees in 1972 who used a tenotomy blade to sever the innervation of the facet joints and successfully treat pain in 998 of his 1000 patients. , After Shealy encountered numerous hematomas with this approach, he used a technique of RFA to lesion the nerve with the aid of fluoroscopic guidance. The initial results of this procedure produced “good” or “excellent” results in 100 of 140 patients with no complications. RFA, or radiofrequency rhizotomy, of a nerve is performed using an insulated needle with an exposed metal tip through which a current is applied. This creates heat and thermal damage to the intended target tissues. In the case of facet-mediated low back pain, the target nerves are medial branches of the dorsal rami that supply the painful facet joints. By lesioning the neural pathway of painful sensory transmission from the painful joints, longer-term pain relief is achieved. This procedure is not curative, but rather provides a palliative treatment for pain; after several months, the nerves reinnervate the facet joint and pain can return. Pain relief following RFA of the medial branch nerve for facet-mediated pain has been reported to extend up to 12 months. This relief is longer than one would expect from typical nerve regrowth after lesioning, which is at a rate of 1 mm per day. It is believed that the heat lesioning by RFA causes a coagulation seal at the lesioning site which requires increased time for repair by endocellular processes, ultimately prolonging symptom improvement. The amount of heat generated to lesion the tissue is a combination of power supplied to the needle and time of lesioning. Two main electrodes are currently marketed to produce the lesion: conventional and water-cooled RFA. Given the more common usage of conventional RFA, the mechanism and procedural techniques discussed further will refer to this design. Subtypes of continuous and pulsed RFA also exist and depend on the pain etiology and targeted nerve. An important technical characteristic of conventional probes is that the burn created by the electrode is in an elliptical pattern with the majority of the lesioning occurring at a maximal distance at the midportion of the shaft of the exposed electrode with very little occurring at the tip. Shape of the RF lesion must be considered during lesioning of the nerve, and in conventional RF the needle should be placed parallel to the nerve rather than perpendicular to it.
Efficacy of RF for Lumbar
With careful patient selection and two successful medical branch blocks, lumbar medial branch RFA can provide substantial long-lasting pain reduction. Dreyfuss (2000) found that 60% of patients reported 90% pain relief and close to 90% of patients achieved at least 60% pain relief at 12 months. A prospective double-blind randomized controlled trial performed by van Kleef et al. (1999) reported improvement in “highest pain” and “mean pain” scores in the RFA treatment group when compared with sham procedure at 8 weeks. Following successful RFA treatment, if pain returns, RFA can be repeated with similar outcomes for pain relief and duration of relief as the initial procedure. ,
Lumbar Radiofrequency Procedure
If using traditional radiofrequency probes, this procedure is ideally done using a parallel approach. As discussed above, the burn area for a traditional probe is in an ovoid shape, without limited lesioning beyond the tip of the probe. As such, in order to maximize nerve ablation, the probe should be placed parallel to the course of the medial branch.
L1–L4 Medial Branch RFA
The patient is placed in prone position, and after sterile prep and drape, an AP view of the level of interest is obtained. Once the target level is centered, the C-arm is obliqued ipsilaterally to visualize the inflection between the superior articular process and the transverse process. Next, progressive caudal tilt of the C-arm is performed until a groove is visualized between the superior articular process and the transverse process. This groove is the target for the radiofrequency probe. This caudal view and subsequent coaxial placement allows for the electrode to approach from an inferolateral angle and ultimately lie in parallel to the medial branch. After a proper view of the target is visualized, the skin is anesthetized. Generally, an 18- or 20-gauge RFA electrode is inserted and advanced in coaxial fashion to the target location aiming for the groove where the caudal aspect of the superior articular process and the superior medial edge of the transverse process meet. Larger diameter needles increase burn area, and gauge should be considered carefully in conjunction with increased needle trauma with larger diameter needles. The needle is advanced so it lies in the groove between the superior articular process and the transverse process, lying parallel to the medial branch nerve. Once ideal placement is obtained, the fluoroscope is rotated to obtain an AP view, in which the active tip should be visualized along the junction of the superior articular process and the transverse process. A lateral view is then obtained to ensure appropriate depth, and to confirm that the active tip is outside of the intervertebral foramen. Lesioning then occurs for 90 s at 80°C.
L5 Dorsal Rami RFA
For the L5 dorsal ramus, the target is the groove between the sacral ala and the superior articular process of the L5–S1 joint. The fluoroscopy arm is rotated ipsilaterally and tilted caudally as the groove at the sacral ala becomes clear. This groove is the target for the radiofrequency probe. This caudal view and subsequent coaxial placement allow for the electrode to approach from an inferolateral angle and ultimately lie in parallel to the dorsal rami. After a proper view of the target is visualized, the skin and subcutaneous tissue along the intended track of the RFA needle is anesthetized. An 18- or 20-gauge RFA electrode is inserted and advanced in coaxial fashion to the target location. The needle is advanced so it lies on the sacral ala, parallel to the L5 dorsal ramus. Once ideal placement is obtained, the fluoroscope is rotated to obtain an AP view and the tilt is removed, and the active tip should be visualized running parallel to the dorsal rami at the sacral ala. A lateral view is then obtained to ensure appropriate depth, and to confirm that the active tip is outside of the intervertebral foramen.
Cervical Radiofrequency Procedure
Correct placement of the electrodes is paramount in order to achieve successful lesioning of the medial branches in the cervical spine and to avoid complications. Depending on which level is being targeted, it is essential to understand and recognize the typical course of each target level as described previously.
The patient is placed in the prone position, and after sterile prep and drape, an initial AP view is obtained to identify target levels. A caudal tilt on fluoroscopy ensures that as the electrodes are placed in coaxial view, they will lie parallel to the direction of the medial branch along the articular pillar. Skin is anesthetized, and an 18- or 20-gauge radiofrequency probe needle is inserted and advanced using intermittent fluoroscopic guidance to maintain needle position parallel to the fluoroscopy beam. The electrode is then advanced in coaxial fashion to the lateral aspect of the articular pillar until os is reached. Once bone is felt and location confirmed with imaging, the RFA needle is redirected laterally to slip past the lateral margin of the articular pillar. The RFA needle is then advanced along the articular pillar while maintaining contact with the lateral aspect of the pillar. A lateral view is then obtained to confirm appropriate depth of the RFA active tip over the target. Further adjustments can be made while using the above technique to ensure that the active tip of the radiofrequency probe is in adequate position. After this is achieved, an AP view is obtained to ensure that the RFA needle is still in contact with the lateral aspect of the pillar.
Stimulation and Lesioning
With precise parallel placement of the radiofrequency probe, the value of sensory and motor stimulation is debated. Motor stimulation can confirm that the needle is not placed in close proximity to nerve roots. During this time, the multifidi should also contract, confirming proper placement. Sensory stimulation can confirm that the probe is stimulating at or near the area of axial pain, and that none is noted into the arm. The probes are then removed, and anesthetic injected before lesioning. Lesioning is then conducted for 90 s at 80°C. If the patient reports any symptoms down the extremity, the lesioning should be immediately discontinued. Although some practitioners choose to inject steroid to avoid possible neuritis, this practice is based primarily on theory, without much evidence to support it.
Epidural Steroid Injections
The cervical spine is composed of seven cervical vertebrae and eight paired cervical nerve roots which supply the upper extremity. C1, which has an anterior and posterior arch but no vertebral body, is commonly referred to as the atlas and C2 as the axis. C2 has a unique feature from other vertebral levels as a bony projection called the dens (or odontoid process) which articulates with C1 allowing for rotation of the head. The remaining cervical vertebrae are overall similar to other vertebrae throughout the spinal column in that they have an anterior vertebral body, posterior spinous process, and paired pedicles, laminae, transverses processes, and articular pillars. The articular pillars are comprised of the pars interarticularis and the superior and inferior articular processes. Between two levels of vertebrae, the superior articular process of the inferior level and the inferior articular process of the superior level form the facet joint on either side. Intervertebral discs separate each of the vertebral bodies. Nerve roots exit the spinal canal through the intervertebral foramen which is comprised in the cervical spine of the pedicles, superior articular process, uncinate process, and vertebral body. In the cervical spine, the intervertebral foramen opens in an oblique, anterolateral direction with the nerve root located posteriorly and inferiorly. The numbering of the cervical nerve roots differs in the cervical spine in comparison to the thoracic and lumbar levels. The cervical nerve roots are named by the vertebral level below the exiting nerve root (i.e., C6 nerve root exits between the C5–C6 intervertebral foramen). The nomenclature differs at the C7–T1 intervertebral foramen where the exiting nerve root is designated C8. Throughout the rest of the spine, the exiting nerve roots derive their name from the vertebral level above (i.e., L4 nerve root exits the L4–L5 intervertebral foramen).
The lumbar spine is composed of five lumbar vertebrae and five paired lumbar nerve roots which supply the lower extremity. The lumbar vertebrae have similar anatomy and have an anterior vertebral body, posterior spinous process, and paired pedicles, laminae, transverses processes, and articular pillars. In the lumbar spine, the nerve root exits the canal posterolaterally via the intervertebral foramen, passing underneath the pedicle. , The lumbar nerve roots are named by the vertebral level above the exiting nerve root (i.e., L4 nerve root exits between the L4–L5 intervertebral foramen).
The spinal cord is covered by three meninges: the innermost pia, arachnoid, and dura mater. The epidural space is a potential space between the ligamentum flavum posteriorly and the dura anteriorly. Posterior to the ligamentum flavum is the interspinous ligament which attaches to the spinous process at the level above and below each level, and the supraspinous ligament which runs along the posterior aspects of the spinous process. These ligaments may be encountered prior to engagement with the ligamentum flavum when performing an epidural injection. The depth of the cervical epidural space, between the ligamentum flavum and the dura, is much smaller than in the lumbar spine. The epidural space at C7–T1 is less than 2–3 mm, whereas the depth of the lumbar epidural space, between the ligamentum flavum and the dura, has been reported to be 4–6 mm. ,
Theoretical Basis of Epidural Injections
Epidural injections are used in the treatment of pain from cervical and lumbar nerve root irritation. Cervical and lumbar radicular pain may occur along the dermatome of the extremity supplied by the nerve root affected. Sensory symptoms can include pain, paresthesias, or numbness. Motor weakness may also be present affecting the muscles innervated by the nerve root involved. Nerve root irritation can occur from a number of different causes: disc herniation, spondylosis, instability, trauma, or tumors. Based on the affected nerve root, symptoms and exam follow a predictable pattern.
Evidence for Cervical Interlaminar Epidural Injections
A prospective study by Castagenera et al. (1994) reported that over 70% of patients with chronic radicular cervical pain obtained at least 75% or greater pain relief at 3, 6, and 12 months following cervical epidural steroid injection. Stav et al. (1993) conducted a randomized prospective study where patients with cervical radicular pain were randomized into receiving local anesthetic with methylprednisolone into the cervical epidural space or cervical muscles. In patients that had the injectate placed into the epidural space, 68% reported greater than 50% pain relief at 1 year from injection compared with less than 12% of patients who had the steroid injected into the cervical muscles. Cicala et al. (1989) found similar results as Stave et al. (1993) at 1-year follow-up. Although promising with their results, the previously stated studies have various shortcomings. None of the studies employed fluoroscopic guidance and all relied solely on the loss of resistance technique to achieve epidural access. Additionally, patient selection in these studies based on symptoms, physical exam, or radiographic imaging was either absent or had a mix of pathophysiology. One prospective study looked at those differences in radicular arm pain at 1, 3, and 6 months between groups randomized to conservative management (gabapentin and/or nortriptyline and physical therapy), cervical interlaminar epidural steroid injection, or combination of both treatments. The only statistically significant positive outcome found by the investigators was at 3 months for those who were in the combination treatment arm.
A single-blind, prospective, randomized, comparative trial by McCormick et al. (2017) evaluated whether using a catheter to deliver steroid to the site of pathology in the cervical spine versus depositing the medication at the site of the standard interlaminar approach (C7–T1) had improved efficacy. Seventy-six patients were randomized, and at 1 month follow-up, the difference between the two groups was not statistically significant. Choi et al. (2016) retrospectively analyzed medical records of 128 patients who underwent fluoroscopically guided cervical epidural steroid injections. Overall, almost 70% of patients had greater than 50% reduction in pain on visual analog scale at second visit (average approximately 1 month later). This was a statistically significant improvement in symptoms across all groups analyzed. The authors also found a statistically significant better response in patients with acute radicular pain secondary to a herniated disc versus patients with chronic neck pain and spinal stenosis.
In a randomized, double-blind, active control trial for patients who had interlaminar epidural injections for cervical disc herniations using lidocaine 0.5% with or without betamethasone, Manchikanti et al. (2013) reported that 72% of patients in the lidocaine group and 68% of patients in the lidocaine and betamethasone group had 50% improvements in pain and function at 2 years.
Evidence for Lumbar Epidural Steroid Injections
Manchikanti et al. (2012) performed a randomized double-blind controlled trial evaluating the effectiveness of interlaminar epidural steroid injections for patients with radicular pain secondary to disc herniations with anesthetic and steroid versus anesthetic alone. Significant improvement was defined as at least 50% improvement with pain relief and functional status. At 1-year follow-up, 67% of patients receiving anesthetic only versus 85% of those who received anesthetic plus steroid reported improvement, with a statistically significant difference. There were also statistically significant improvements found at 3- and 6-month follow-up in the steroid treatment group. Manchikanti et al. (2014) continued to follow these patients and published their findings. At 2 years, they found significant improvement in 70% of patient who received the steroid and anesthetic versus 60% who received the anesthetic alone. In a 2011 study by Kim and Brown, patients with lumbar radicular pain were randomized to receive anesthetic mixed with equipotent doses of either dexamethasone or methylprednisolone. The VAS scores were documented at an average follow-up day of 51.1 for the methylprednisolone group and 41.1 for the dexamethasone group, which was not statistically significant. Decreases in VAS from preprocedure to follow-up were 19.7% and 27.2% for dexamethasone and methylprednisolone, respectively, but the differences were not statistically significant.
Cervical Epidural Injection Procedure
The patient is placed in prone position, and after sterile prep and drape, an AP view of the level of interest is obtained. Once the target level is centered, typically the C7–T1 interlaminar space, the endplates of the vertebrae above and below are squared off. The initial target should be the superior aspect of the lamina at the level below the interlaminar space of intended entry, just lateral to the spinous process. Skin is anesthetized and the needle is then advanced to the superior aspect of the inferior lamina until os is reached. Targeting os of the lamina allows safe advancement of the needle. Once os is reached, the needle is then redirected superiorly and advanced until the ligamentum flavum is engaged. Next, loss of resistance technique is used until the epidural space is entered.
A contralateral oblique image should then be obtained to check depth. Lateral views can be difficult given the patient’s shoulders may obstruct fluoroscopic visualization, so a contralateral oblique is recommended. The fluoroscope is typically obliqued 45–55° contralateral to needle tip to clearly view the laminar line. Once the needle tip approaches the ventral aspect of the lamina, one can start to expect loss of resistance.
Lumbar Interlaminar Epidural Steroid Injection Procedure
The patient is placed in prone positioning with a pillow placed under the abdomen to reduce lumbar lordosis and help open the interlaminar space for easier visualization and entry. After sterile prep and drape, an AP view of the level of interest is obtained. Once the target level is centered, the superior endplate of the vertebrae below the target space and the inferior endplate of the vertebrae above are squared off. In this view, the interlaminar space should be maximally accessible. Given the variation in patient anatomy, the fluoroscope may need to be tilted either cephalad or caudad to increase accessibility of the interlaminar space. The initial target should be the superior aspect of the lamina at the level below the interlaminar space of intended entry, just lateral to the spinous process. The final needle target may be paramedian or directed midline. After proper identification, the skin is anesthetized. Then a Tuohy needle is advanced to the superior aspect of the inferior lamina until os is reached. Targeting the lamina helps determine depth and provides a firm endpoint. Once os is reached, the needle is then redirected superiorly and advanced until the ligamentum flavum is engaged. Next, loss of resistance technique is used until the epidural space is entered.
Depending on preference, either a lateral or contralateral oblique image can be obtained to check depth. If a lateral is used, as with the cervical spine, knowledge of the lumbar lamina is necessary to understand how the imaging relates to the true needle depth. For an accurate view of needle depth, the needle must be lateral to midline and extra care should be taken to obtain a contralateral, rather than an ipsilateral, view otherwise the image will not show appropriate depth. Once the needle tip approaches the ventral aspect of the laminar line, one can start to expect loss of resistance.
Loss of Resistance Technique for Interlaminar Epidural Steroid Injections
A loss of resistance technique is used to determine when the needle tip has entered the epidural space. This technique uses a loss of resistance syringe which can be plastic or glass, using air, saline, or both. The loss of resistance technique is performed by advancing a Tuohy needle through the skin, subcutaneous tissue, and posterior muscles. When the ligamentum flavum is reached, the stylet is removed from the needle and the loss of resistance syringe is attached. The ligament provides resistance and prevents the plunger from advancing in the syringe. The needle is then advanced slowly while pressure is applied to the plunger. When the needle tip enters the epidural space, a loss of resistance will be felt. If loss of resistance begins prior to the needle being at the level of the ligament, a false loss is the likely source. A false loss can also occur if the needle tip is advanced too far and goes through the dura into the subdural space. Care should be taken not to advance the needle past the epidural space to avoid potential complications of neural injury, and post dural puncture headache.
After loss of resistance is achieved, the syringe is removed from the needle, with observation for cerebrospinal fluid or blood flow from the needle. If CSF or blood flow is encountered, the procedure should be abandoned at that location and the procedure should be rescheduled or another location should be attempted. If no CSF or blood is encountered, extension tubing is attached to the Tuohy needle and 0.5–1 mL of contrast is injected under live fluoroscopy to confirm the needle tip is not placed intravascularly and epidural spread is seen. After appropriate epidural contrast spread is confirmed, nonparticulate steroid is injected, and the needle is removed.
Lumbar Transforaminal Epidural Steroid Injections
The lumbar intervertebral foramen is formed between the pedicles of the vertebrae superiorly and inferiorly. The lamina and facet joint form the posterior border and the vertebral bodies and intervertebral disc anteriorly. In the lumbar spine, the intervertebral foramen is directed laterally with the nerve root exiting under the superior pedicle. The nerve roots enter the intervertebral foramen from the spinal cord along the medial and inferior aspect of the superior pedicle and travel obliquely, inferiorly, and laterally. Reinforcing medullary arteries traverse the intervertebral foramen along with the nerve root. The largest and most clinically relevant of these is the artery of Adamkiewicz (arteria radicularis magna). This artery typically arises from the T5 to L5 levels and is on the left side in 69%–85% of the population.
Theoretical Basis of Transforaminal Epidural Injections
Lumbar transforaminal epidural steroid injections are used for the treatment of lumbar radicular pain. Patient selection for transforaminal steroid injections is similar to that for interlaminar epidural steroid injections; however, this procedure should generally be done with unilateral symptoms. The advantage of the transforaminal approach compared with interlaminar or caudal techniques is that the transforaminal route target is more specific to the intended nerve root. In addition, it has the advantage that the injectate reaches the ventrolateral epidural space. The disadvantage of the transforaminal approach, when compared with interlaminar, is that it conveys more risk, particularly of arterial injection and subsequent paraplegia. This is frequently thought to occur due to intraarterial injection into the artery of Adamkiewicz. Additionally, particulate steroid preparations are thought to convey higher risk of spinal cord infarction than nonparticulate steroids.
Vad et al. (2002) conducted a prospective, randomized study with participants receiving either a transforaminal epidural steroid injection or a saline trigger-point injection. Successful outcomes were determined by subjective report of “good” or “very good” improvement and a 50% reduction in their pain. At an average follow-up of 1.4 years, the transforaminal epidural group had a statistically significant success rate of 84% compared with 48% of the saline trigger-point group. Ackerman and Ahmad (2007) compared different approaches to the epidural space in three separate groups: caudal, interlaminar, or transforaminal. Ninety participants were randomly assigned to each group and were evaluated at 24 weeks by reporting “complete relief,” “partial relief,” or “no relief.” At 24 weeks, 25 of the 30 patients who underwent transforaminal epidural steroid injection reported either “complete relief” or “partial relief” which was statistically significant when compared with the other two approaches. A prospective, randomized study comparing interlaminar and transforaminal epidural steroid injections for chronic unilateral radicular pain by Rados et al. (2010) found that 63% of patients in the transforaminal group achieved at least 50% pain relief at 24 weeks compared with 53% in the interlaminar group. They did not find these differences to be statistically significant. Jeong et al. (2007) evaluated the effectiveness of injections at the level of disc herniation (preganglionic) affected the nerve root or at the level where the nerve root exited the foramen (ganglionic) at 3 and 6 months follow-up. At 3 months follow-up, 88% of patients reported subjective pain relief with injection at the level of the disc herniation and 71% reported improvement with injection at the level of the nerve root exiting. At 6 months, no statistical significance was found between the two groups with 60% reporting “good” or “excellent” outcomes in the preganglionic group versus 67% in the ganglionic group. A similar study by Lee et al. (2006) found that 90% of patients who received a preganglionic epidural steroid injection and 69% of patients receiving a ganglionic injection after 2 weeks obtained a greater than 50% reduction of pain by visual assessment scale. In a randomized, double-blind controlled trial by Karppinen (2001) comparing transforaminal injection of either steroid or saline found only a favorable difference between the two groups at 2 weeks follow-up. At 4 weeks, there was no significant difference between the two groups. At 3- and 6-month follow-up, there was a statistically significant effect in favor of the saline group. At 1 year, no statistical significance between the two groups was found. This study has been often criticized for not using a true placebo as the control group had an injection of saline. Ghahreman et al. (2010) designed a prospective, randomized study of 150 patients comparing the efficacy of transforaminal steroid injection or intramuscular erector spinae muscle injection using local anesthetic, local anesthetic with steroid or normal saline. At 1 month, the number of responders (at least 50% reduction of radicular pain) in the group who received transforaminal epidural steroid injections (54%) was significantly greater than the other groups. Patients who responded to transforaminal normal saline (19%), transforaminal local anesthetic (7%), intramuscular steroids (18%), or intramuscular saline (13%) did not have a significant difference between groups.
Procedure for Transforaminal Epidural Steroid Injections
The patient is placed in prone position and after sterile prep and drape, AP view of the target level is obtained. The superior endplate of the vertebra of the intended foramen is aligned. In order to optimally enter the foramen, the C-arm should be obliqued ipsilaterally. The final needle tip position below the inferior and midline portion of the pedicle, or the six o’clock position. The “safe triangle” is formed by the borders of the exiting nerve root, the inferior border of the pedicle, and an imaginary sagittal line from the lateral aspect of the pedicle. Of note, the safety triangle is generally considered safe as it reduces the chance of an inadvertent nerve injury. It does not, however, prevent potential injury to blood vessels in the vicinity of the nerve root.
A 22-gauge 3 ½- or 5-inch needle is then inserted and advanced to the inferior border at the six o’clock position of the pedicle using multiplanar fluoroscopy. The needle is then redirected inferiorly and advanced into the foramen ( Fig. 13.4 ). An AP view is then used to confirm needle position ( Fig. 13.5 ). Next, a lateral view is obtained to verify the depth of the needle in the foramen. Correct placement should visualize the needle tip just inside the foramen ( Fig. 13.6 ). The C-arm is then positioned in an AP view, and under continuous fluoroscopy, contrast is injected through extension tubing to confirm epidural spread without vascular or intrathecal uptake ( Fig. 13.7 ). Steroid and often local anesthetic are then injected and the needle is withdrawn.