Epidural steroid injections (ESIs) for the management of lumbar radicular pain provide the advantage of delivering potent anti-inflammatory agents in a localized fashion to the area of affected nerve roots, thereby decreasing the systemic side effects often seen with orally administered steroids. Lumbar ESIs have been endorsed by the North American Spine Society (2) and the Agency for Health Care Policy and Research (3) as an integral part of nonsurgical management of radicular pain from lumbar spine disorders. Cervical and thoracic ESIs are also being used to treat cervical and thoracic radicular pain respectively, but have not, as of the time of this writing, received similar endorsement.
Pathophysiology of Radicular Spinal Pain
Radicular pain is the result of nerve root irritation or inflammation (4–13). It is often described as a sharp, lancinating, and cramping discomfort that may feel like it is “shooting” from the spine along a dermatomal distribution. Other clinical manifestations of nerve root inflammation may include dermatomal hypesthesia or hyperalgesia, weakness of muscle groups innervated by the involved nerve roots, diminished deep tendon reflexes, and positive neural tension signs such as the straight-leg-raising test or Spurling’s maneuver. Animal research in rats has revealed severe local inflammation in the epidural space and nerve root after injection of autologous nuclear material into the epidural space (6). High levels of PLA2, an enzyme that regulates the initial inflammatory cascade, have been demonstrated in herniated disc material from surgical samples in humans. Leukotriene B4, thromboxane B2, and inflammatory products also have been discovered within herniated human discs after surgery (4). Animal models have demonstrated that injection of PLA2 into the epidural space induces local demyelination of nerve roots and subsequent ectopic nerve discharges, which is considered to be the primary pathophysiologic mechanism of radicular pain (5). Interleukin 1 and tumor necrosis factor-a (TNF-α) have been detected in the herniated nucleus pulposus and found to play an important role in the pathogenesis of radicular pain from a herniated nucleus pulposus (11,12,14,14a).
Radicular low back pain caused by spinal stenosis probably occurs through the impedance of normal nerve root vascular flow and subsequent development of nerve root malnutrition, nerve root edema, and nerve root dysfunction (13). Chronic nerve root compression can induce axonal ischemia, impede venous return, promote extravasation of plasma proteins, and cause local inflammation (7). Venous congestion and arterial compromise can induce radiculopathy.
In a rat study, the compression of dorsal root ganglion (DRG) resulted in the reduction in K+, a hyperpolarizing shift in TTX-S Na+ current activation, and an enhanced TTX-R Na+ current. These phenomena may all contribute to the enhanced neuronal excitability and thus to the pain and hyperalgesia associated with the compression of DRG. In a separate study, an inflammatory soup (IS) consisting of bradykinin, serotonin, prostaglandin E2, and histamine (each 10−6 M) was applied topically to the DRG that had undergone chronic compression. IS remarkably increased the discharge rates of somata of a chronically compressed dorsal root ganglion (CCD) in rat neurons and evoked discharges in more silent-CCD than control neurons. Inflammatory mediators, by increasing the excitability of DRG somata, may contribute to chronic compressioninduced neuronal hyperexcitability and to hyperalgesia and tactile allodynia (15). That these findings of lumbar radicular pain may be associated with increased excitability of involved DRG neurons was confirmed by a similar animal study (16).
In summary, clinical practice and animal research suggest that radicular spinal pain is the result of nerve root inflammation in the epidural space that is provoked by leakage of disc material, compression of the nerve root, and/or irritation of DRG from spinal stenosis. In addition, the pain may also relate to the enhanced neuronal excitability of the DRGs associated with chronic compression that occurs in neuroforaminal stenosis due to spondylosis or nucleus pulposus herniation.
Because radicular pain appears to originate from inflammation within the epidural space and nerve root, analgesic effects of epidural corticosteroids most likely are related to their anti-inflammatory effect. Underlying corticosteroid anti-inflammatory mechanisms include inhibition of phospholipase A2 and inflammation (14, 14a), inhibition of neural transmission in nociceptive C fibers (5), reduction of capillary permeability, and nonselective inhibition of TNF-α (12) and IL-1 (17).
Rationale for Epidural Injection of Corticosteroids and Local Anesthetics
Given the fact that the efficacy of corticosteroids apparently depends on their anti-inflammatory mechanism, it stands to reason that epidural injection should have higher efficacy in reducing nociceptive-induced radicular pain if applied earlier in the inflammatory process. In contrast, at least some of the nerve root fibers undergo fibrosis and necrosis in chronic radicular pain states. This perhaps explains in part why corticosteroid injections are less beneficial in the chronic setting. In fact, one study demonstrated little efficacy of cervical ESIs in atraumatic (i.e., stretched or compressed nerve roots) neuropathic radicular pain (68). In addition to timing of administration, the actual route of administration is important in treating radicular pain. For example, one study found that corticosteroid administration through the intravenous approach only offered short-term (<3 days) benefit (18). However, an open trial without control patients showed that intravenous methylprednisolone resulted in improvement in all 11 patients at the median follow-up time of 3.8 months (19). At the time of this writing, there have been only a couple case reports and no peer-reviewed, randomized controlled studies on the benefit of oral steroids in the treatment of radiculopathy. Some case reports provide initial data (20,21).
Local anesthetics are often administered along with corticosteroids during epidural injections. By blocking sodium channels, local anesthetics impair peripheral neurotransmission of pain impulses, normalize the hyperalgesic state of the nervous system, and prevent and/or reduce the neuronal plasticity in the central nervous system by reducing the peripheral nociceptive input. Perhaps this, in part, explains the well-recognized phenomenon in clinical practice of pain relief after injection of local anesthetic often outlasting the physiological action of the anesthetics. In chronic (>6 months in duration) radiculopathy, neuropathic pain likely plays a greater role than pain due to inflammation. Since local anesthetics act directly on axons rather than acting as anti-inflammatory agents like corticosteroids, one would anticipate that they have efficacy in chronic radiculopathy. In fact, one study demonstrated the important role of local anesthetics in reducing chronic cervical radicular pain and discussed their possible mechanism via the neuronal plasticity that has been proposed to play a role in chronic radiculopathy (22).
Indications for ESIs
The primary indication for ESIs is radicular pain associated with a herniated nucleus pulposus or spinal stenosis. A variety of other indications have been reported with variable results (23,25–29). These include radicular pain associated with lumbar spine compression fracture, facet or nerve root cysts, postlaminectomy back pain, cervical strain syndromes with associated myofascial pain, and postherpetic neuralgia (23,25–29).
Contraindications for ESIs
Contraindications for epidural corticosteroid injections include systemic infection, local infection at the site of planned injection, bleeding disorder or full anticoagulation, history of significant allergic reactions to the components of the solution for injection, severe central canal stenosis at the level of planned injection, and lumbar ESI in pregnant women (23,25–29). Caution should be used when performing injections in patients with poorly controlled diabetes and in individuals who have a history of severe or uncontrolled hypertension or congestive heart failure (CHF), because of the potential for steroid-induced fluid retention.
Cervical, thoracic, and lumbar epidural injections may be performed through either interlaminar or transforaminal approaches, and lumbosacral injections may also be performed through the caudal route (23,25–29).
Timing, Frequency, Dose, and Volume of Epidural Injections
Optimal timing of ESIs is unknown, although there is evidence of better benefit if ESIs are performed within 3 months of radicular pain onset (36,37). The general consensus is that most patients with radicular symptoms should undergo a few weeks of treatment including oral medications, physical therapy or manual medicine, and relative rest from activities that exacerbate their pain, before undergoing ESIs (3,23,26). If a patient does not have success with such a program, or if the therapy cannot progress because the patient’s pain is too severe, an ESI is indicated for pain control. In contrast, ESIs can be considered earlier in patients with severe radicular pain not responding to even opioid medication or with pain that is significantly interfering with a patient’s sleep and/or function (26,28). Early ESIs also carry the theoretical benefit of controlling inflammation at an early stage (5,7,38) and possibly preventing permanent neural damage such as nerve fibrosis from prolonged inflammation (8). A study demonstrated that an ESI has higher efficacy (>75% pain relief ) for patients with radicular pain within 3 months duration, whereas less benefit was found in patients with sciatica longer than 7 months (39). Another study compared epidural injections with bupivacaine alone versus injections of bupivacaine with methylprednisolone in patients with lumbar radicular pain longer than 6 months in duration. At 3-month follow-up, both treatments reduced pain but there was no additional benefit with corticosteroids (40).
The time interval between epidural injections should vary depending upon the steroid preparation used. Because injected methylprednisolone is reported to remain in situ for about 2 weeks (41), the clinician should probably consider waiting for about 2 weeks before fully assessing a patient’s response or administering a repeat injection.
Studies have suggested that the total maximum methylprednisolone dose should be about 3 mg/kg of body weight because excessive salt and water retention can occur at doses above this due to the mineralocorticoid properties of corticosteroids. In general, it is felt that up to three to four ESIs within a year may be performed if clinically indicated (23). Some clinicians schedule and proceed with a series of three ESIs regardless of the clinical response to the first preceding injection(s). Although the efficacy of this approach is unclear, as there are no medical outcome studies to support or refute such a regimen, it may be best to reassess the response to a given injection at the time of an intervening office visit before proceeding with another injection. Using this approach, the clinician can determine if another injection is still needed and can more readily alter their planned injection technique, rather than trying to make this assessment at the time of the scheduled injection itself.
Recommended injection volumes and the corticosteroid doses are dictated mainly by the approach used as shown in Tables 68-1 and 68-2 (23,25,26,29). The epidural steroid can be injected in a preservative-free diluent such as lidocaine (1% to 2%) or normal saline (25,28,29). In the cervical spine, it is recommended that local anesthetics and steroids be injected separately to prevent a potential embolus of poorly dissolved steroid particles within a local anesthetic diluent. A recent study demonstrated that a nonparticulate dexamethasone has similar efficacy compared with particulate triamcinolone, and carries the lowest potential risk of embolization with inadvertent intravascular injection when used in ESIs (41).