Version July 2025
INTRODUCTION —
Up to 84 percent of adults will experience an episode of low back pain at least once [1,2]. For most patients, low back pain is self-limited and with a favorable long-term prognosis [3,4]. For a subset of patients, back pain symptoms may persist despite conservative management. Though meaningful improvement is still possible, the likelihood of full recovery declines over time.
Patients with low back pain are categorized by the duration of symptoms, which guides candidacy for treatment:
●Acute low back pain is defined by a duration of less than four weeks.
●Subacute low back pain represents a transition period between acute and chronic back pain in which symptoms are present for more than 4 and less than 14 weeks.
●Chronic low back pain is defined as pain that persists for 12 or more weeks.
Most acute low back pain resolves with conservative treatment. Most subacute and chronic low back pain is managed with nonpharmacologic and pharmacologic treatment. Patients with persistent pain despite such treatment for at least six weeks may be considered for nonsurgical interventional treatment. Typically, these treatments are offered for patients with radicular features, symptoms attributed to radiologic evidence of spinal degenerative changes, and/or positive responses to selected diagnostic or prognostic testing as clinical trials have focused primarily on these populations. However, in clinical practice, nonsurgical interventional treatments are sometimes offered to patients with refractory, nonspecific pain without these features. Surgery is reserved for patients with severe or progressive symptoms, weakness, or other neurologic deficits.
This topic will address nonsurgical interventional treatment for patients with subacute or chronic low back pain. Management of acute low back pain and other treatment modalities for subacute and chronic low back pain are discussed separately:
●(See "Treatment of acute low back pain".)
●(See "Subacute and chronic low back pain: Management".)
●(See "Subacute and chronic low back pain: Surgical treatment".)
CONFIRM CANDIDACY FOR NONSURGICAL INTERVENTIONAL TREATMENT —
Patients with subacute and chronic low back pain should undergo the following evaluation prior to consideration for interventional treatment:
●Identify indications for urgent neurologic or surgical referral – Patients with features of cauda equina syndrome, including bilateral radicular signs or symptoms, urinary retention, or saddle anesthesia, should be referred for urgent neuroimaging and escalation of care (algorithm 1). The management of acute radiculopathy and lumbar spinal stenosis are discussed in separate topic reviews. (See "Acute lumbosacral radiculopathy: Treatment and prognosis", section on 'Identify patients with high risk mechanisms' and "Lumbar spinal stenosis: Treatment and prognosis".)
●Identify and address specific etiologies of low back pain – Patients should undergo a history and physical examination at the initial presentation of low back pain to a health care provider, regardless of back pain duration. This initial evaluation assists with identifying specific etiologies of low back pain and features that warrant additional workup, reserving imaging for selected cases with risk factors for those conditions (table 1 and table 2) (see "Evaluation of low back pain in adults", section on 'Initial evaluation' and "Evaluation of low back pain in adults", section on 'Diagnostic imaging'). If specific etiologies are identified, management is focused on addressing the underlying condition.
When an underlying etiology is not identified, the patient is given a presumptive diagnosis of "nonspecific low back pain." Initial management is with conservative therapies, as below.
●Complete a trial of conservative therapy – In the absence of serious pathology or severe neurologic deficits, a trial of at least six weeks of conservative treatment is indicated for all patients before interventional treatments are pursued. This may include exercise or physical therapy, nonopioid analgesic medications, and other therapies. (See "Subacute and chronic low back pain: Management".)
DETERMINE IF THE PAIN IS RADICULAR VERSUS NONRADICULAR —
For patients with low back pain that persists beyond six weeks of conservative therapy, we determine candidacy for interventional procedures based on the presence or absence of radicular features:
●Radicular pain – Radicular pain originating in the lower back with radiation down the leg in a dermatomal or nerve root distribution is suggestive of lumbosacral radiculopathy. Radicular pain may be accompanied by weakness, reflex changes, or dermatomal sensory loss (table 3), though these additional findings are not always clinically apparent. Most commonly, symptoms are secondary to a herniated disc or spinal stenosis from degenerative changes. Patients with subacute or chronic low back pain with radicular features are candidates for epidural glucocorticoid injections and radiofrequency interventions, as below. (See 'Radicular pain' below.)
Details regarding the diagnosis of lumbosacral radiculopathy are discussed separately. (See "Acute lumbosacral radiculopathy: Etiology, clinical features, and diagnosis", section on 'Evaluation and diagnosis'.)
●Nonradicular pain – Most cases of low back pain without radicular features or a specific underlying condition are self-limited; however, some patients may go on to develop subacute or chronic back pain that persists beyond 12 weeks despite conservative management with nonpharmacologic and pharmacologic treatment. A subset of these patients may be candidates for nonsurgical interventional treatments, as identified by a combination of clinical evaluation, imaging, and/or interventional diagnostic procedures (see 'Nonradicular pain' below and 'Pretreatment evaluation' below). Treatment options for patients with nonradicular, nonspecific low back pain who are not candidates for nonsurgical interventional treatments are discussed in further detail separately. (See 'Patients without indication for interventional treatment' below and "Approach to the management of chronic non-cancer pain in adults".)
RADICULAR PAIN
Pretreatment evaluation
●Imaging – Initial assessment with magnetic resonance imaging (MRI; preferred) or computed tomography (CT) is typically performed to evaluate lumbosacral radiculopathy (algorithm 1). Imaging can identify patients with a herniated disc or other pathology that may be responsive to epidural glucocorticoid injections. Though a randomized trial found no difference in outcomes between individuals with sciatica randomized to receive an MRI before an epidural steroid injection and those who had an MRI that was not considered by the treating clinician, patients in whom MRI data were not considered had worse outcomes if the injection performed differed from that proposed by an independent reviewer [5]. (See "Acute lumbosacral radiculopathy: Etiology, clinical features, and diagnosis", section on 'Neuroimaging'.)
●Electromyography (EMG)/nerve conduction study (NCS) – EMG/NCS may be pursued to evaluate weakness due to radiculopathy but is not required prior to glucocorticoid injection. (See "Acute lumbosacral radiculopathy: Etiology, clinical features, and diagnosis", section on 'Neurodiagnostic testing'.)
●Interventional diagnostic procedures – Selective nerve root blocks are sometimes used for diagnostic evaluation and to predict response to spinal decompression procedures. Some studies have also reported intermediate-term pain relief and reduced need for operative intervention [6,7]. In practice, however, the injectate can spread to contiguous levels via the epidural space, which may misdiagnose the level of pathology and interfere with spinal blood supply [8-10]. As a result, these procedures also carry a small risk of spinal cord injury, which must be weighed against the potential benefit of diagnostic information and/or avoiding operative intervention altogether [10-12].
Interventional diagnostic procedures for chronic low back pain are discussed in further detail below. (See 'Procedure' below.)
Interventional treatment for radicular pain
Epidural glucocorticoid injections
Indications and contraindications — We offer epidural glucocorticoid injections (also referred to as "corticosteroid injections," "steroid injections," or "ESI") for patients with lumbosacral radiculopathy that has persisted despite at least six weeks of conservative treatment [13-15]. In some cases, interventional treatment is also offered for patients with chronic, refractory back pain symptoms who decline surgery or are not candidates for surgery. Evidence supporting their use is mixed but generally suggests modest benefit. As examples:
●Herniated disc – In trials of epidural glucocorticoid injection for patients with radiculopathy due to a herniated disc, there is short-term, but not long-term, improvement in pain; the average improvement in pain intensity is generally modest [16,17]. In a systematic review of randomized trials, epidural corticosteroid injection was associated with a small improvement in leg pain (mean difference [MD] 7.6 on a pain scale of 0 to 100, 95% CI 3.7-11.4) and disability (standardized MD 0.33, 95% CI 0.09-0.56) at two weeks and decreased risk of surgery at up to three months (relative risk 0.62, 95% CI 0.41-0.92) [17]. The trials included heterogeneous groups of patients with variable duration of symptoms; treatments varied considerably among trials, and effects were below the predefined thresholds of clinical importance. Two subsequent systematic reviews reported similar results [18,19].
●Spinal stenosis – The evidence supporting epidural glucocorticoid injections for underlying spinal stenosis is less clear [20] as randomized trials have provided inconsistent results and are also limited by heterogeneous populations and interventions. These are reviewed in further detail separately. (See "Lumbar spinal stenosis: Treatment and prognosis", section on 'Epidural injections'.)
Despite the relatively small benefit seen in clinical trials [15], our clinical experience suggests that some patients with radiculopathy from spinal stenosis may obtain meaningful relief from epidural corticosteroid injection, particularly in patients who wish to avoid surgery or who have contraindications to surgery. Because the evidence for benefit in spinal stenosis from other etiologies is less robust than for herniated disc, providers should carefully weigh the potential benefits against the risk of serious complications, including dural puncture, infarct, and vertebral fracture [16,21,22]. The management of lumbar spinal stenosis is reviewed in further detail separately. (See "Lumbar spinal stenosis: Treatment and prognosis".)
Epidural glucocorticoid injections are contraindicated in patients with infection at the injection site, systemic infection, or coagulopathy. Osteoporosis is a relative contraindication that requires a shared decision-making approach regarding risk and benefit with the patient. (See 'Adverse effects' below.)
Procedure
●Initial injection – Interlaminar epidural glucocorticoid injections involve the administration of steroids via a needle inserted in the epidural space between the ligamentum flavum and the dura (figure 1). Injections are administered by various spine specialists, including anesthesiologists, physiatrists, and interventional radiologists. The procedure is typically performed prone, with fluoroscopy, and with local anesthesia. The use of sedation varies among institutions, ranging from no sedation to deep sedation, and may include the use of monitored anesthesia care. Guidelines do not recommend the routine use of sedation, as studies have not found it necessary and deep sedation can increase the chance of a complication [23].
Epidural injections can be performed by the interlaminar approach (via the interlaminar space in the spine as noted above) and the transforaminal approach (through the neuroforamen dorsal to the nerve root). The caudal approach (through the sacral hiatus at the sacral canal) may be useful in patients with prior surgery and for patients on anticoagulation as discontinuation of anticoagulation is not routinely required (figure 2). If fluoroscopy is used, contrast dye is usually injected to confirm the placement of the needle and injectate in the epidural space.
Studies suggest better outcomes and fewer side effects with the following:
•Injecting at the level of radiologic finding rather than the level of the dermatomal distribution (eg, between L4 and L5 for a herniated disc causing L5 symptoms) [24,25]
•Using a transforaminal approach in those with unilateral pain [26,27]
•Adding local anesthetic to the injection mixture [25,26]
•Using particulate (long-acting) rather than nonparticulate or soluble steroids for all interlaminar and lower lumbar transforaminal injections [24,28]
•Not exceeding an equivalent dose of 40 mg of methylprednisolone with any single injection to avoid the cumulative effects of steroids [24,25,29]
●Postprocedure guidance – We inform patients that they may not appreciate the benefits of the epidural glucocorticoid injection for at least 24 to 72 hours. Response varies as some patients may feel immediate relief (particularly if a local anesthetic is used) and others may not feel substantial benefits until days or weeks later. This variation may be because the injected glucocorticoid has both direct effects in the spine and systemic effects.
As an outpatient/ambulatory procedure, the patient can usually go home after a brief period of postprocedure observation. We follow up at approximately four weeks and consider additional injections, as below.
Repeat injections — Consistent with guidelines, the number and type of injections should be tailored to a patient's response to the initial injection [24]. Once symptom resolution is achieved, additional injections to complete a predetermined "series" are not indicated. No more than four injections should be given at the same site within 12 months because of concerns about potential suppression of the hypothalamic-pituitary-adrenal axis and the cumulative effects of steroids (eg, increased fracture risk), though there is little published evidence to support these parameters.
●Patients with a suboptimal response – In a patient with persistent symptoms after the initial injection, a modified injection approach (eg, transforaminal instead of interlaminar), steroid preparation (eg, a particulate instead of nonparticulate steroid), or spinal level may yield an improved subsequent result. If symptoms persist despite a modified approach, we do not perform additional repeat injections and instead pursue alternative treatment options. (See 'Radiofrequency interventions for refractory symptoms' below.)
●Patients with symptom recurrence after initial response – Repeat injections may be performed if symptoms recur, usually no more frequently than every three months.
Adverse effects — Potential benefits must be weighed against the risk of rare but serious adverse effects. A systematic review found that methods for assessing the harms of epidural glucocorticoid injections in randomized trials were not well reported and harm data were sparse [17]. Thirteen of 30 placebo-controlled trials did not report harm data or reported no harm. The limited data report the following:
●Minor adverse events – Large observational studies have found that minor adverse events (eg, bleeding, dural puncture with cerebrospinal fluid leak, transient nerve root irritation) occur in <1 percent of procedures [30,31]. Transient systemic glucocorticoid-related effects (eg, significant blood sugar elevation, gastrointestinal or psychologic symptoms, vertigo, or dizziness) were observed in 0.1 percent of procedures.
●Compression fracture – Patients considering a trial of epidural glucocorticoids should be made aware of a possible small increase in vertebral fracture risk associated with each injection due to the increased risk of bone fragility. In a large retrospective database analysis, each successive epidural glucocorticoid injection increased the five-year risk of vertebral body fracture compared with a matched cohort of noninjected patients by a factor of 1.21 (95% CI 1.08-1.30) after adjustment of covariates [32]. The risk of compression fracture may be mitigated somewhat by antiosteoporotic medication during the treatment course [21].
●Major complications – Major complications are rare. In 2014, the US Food and Drug Administration issued a drug safety communication about epidural injection of glucocorticoids, noting the potential for rare but serious adverse effects (stroke, paralysis, death, and loss of vision due to the rapid rise of retinal venous pressure when large volumes are injected rapidly) [33,34]. Paralysis following lumbar transforaminal injections has been attributed to spinal cord infarct stemming from either injury, spasm, or embolization of particulate steroids injected into a radiculomedullary artery feeding the spinal cord [35]. In a systematic review of randomized trials (2912 patients in total), one serious adverse event (a case of retroperitoneal hematoma in a patient receiving anticoagulation) was reported in one trial [36]. In large observational studies, major complications resulting in permanent neurologic sequelae (eg, spine hematoma, infection) were similarly rare, with an incidence of 0.01 percent or lower [31,37].
●Contaminated medication – In 2012, hundreds of patients developed fungal meningitis, in many cases fatal, from epidural injections of contaminated glucocorticoid from a single compounding pharmacy [38]. (See "Central nervous system infections due to dematiaceous fungi (cerebral phaeohyphomycosis)", section on 'Outbreak of fungal meningitis and osteoarticular infections'.)
Radiofrequency interventions for refractory symptoms
●Pulsed radiofrequency – Pulsed radiofrequency uses intermittent radio waves to create an electrical field surrounding neural tissue at nonablative temperatures.
Pulsed radiofrequency of the dorsal root ganglia is a treatment option for patients with chronic low back pain with radicular features who have not responded to epidural corticosteroid injection. It is also an alternative treatment option for patients with osteoporosis or other contraindications to corticosteroid injections. This intervention is usually considered alongside surgical options (eg, discectomy) using a shared decision-making approach.
In a meta-analysis of 10 randomized trials, pulsed radiofrequency of the lumbar dorsal root ganglion modestly improved pain scores at three months when compared with epidural injection (MD -1.31, 95% CI -2.29 to 0.33); however, heterogeneity of patient selection, pulsed radiofrequency technique, and epidural injection content limit generalizability [39]. Furthermore, this improvement was not sustained at six months, and there was no difference in average disability at any time point. There are no guidelines available for how often pulsed radiofrequency can be repeated, though it is usually offered no more than twice per year.
Adverse effects from pulsed radiofrequency can include pain at the procedure site or down the leg if a nerve root is injured, bleeding, infection, and equipment-related complications (eg, skin burns).
●Radiofrequency ablation (RFA) – RFA (eg, nonpulsed) has not been effective in the treatment of chronic low back pain with radicular symptoms [40]. We do not use conventional (heat) RFA in the treatment of chronic, radicular pain, as studies do not demonstrate improved outcomes over control treatment.
The use of RFA in the treatment of chronic, nonradicular pain is discussed separately. (See 'Radiofrequency ablation' below and "Interventional therapies for chronic pain", section on 'Radiofrequency neurotomy (radiofrequency ablation)'.)
Other procedures for spinal stenosis — Patients with spinal stenosis may be candidates for additional procedures, including minimally invasive lumbar decompression and intraspinous spacer implantation. These procedures are performed by both surgeons and nonsurgeon interventionalists. A detailed discussion of these procedures is presented in further detail separately. (See "Lumbar spinal stenosis: Treatment and prognosis", section on 'Minimally invasive decompression' and "Lumbar spinal stenosis: Treatment and prognosis", section on 'Intraspinous spacer implantation'.)
NONRADICULAR PAIN
Pretreatment evaluation — Patients with nonspecific, nonradicular pain are evaluated for nonsurgical interventional treatment through a combination of clinical evaluation, MRI, and/or interventional diagnostic procedures.
MRI — MRI is useful in supporting the following conditions that may respond to specific interventional treatments, as below:
●Discogenic or vertebrogenic pain – In trials evaluating interventional treatment for patients with chronic, nonspecific low back pain, the radiologic findings on MRI are used as a marker for discogenic and vertebrogenic disease. These findings are commonly called "Modic changes," classified as types 1 to 3 depending on the extent of inflammation and marrow changes [41]. While Modic changes are presumed to be related to symptoms, their significance remains uncertain as they have also been noted in asymptomatic patients. (See "Evaluation of low back pain in adults", section on 'Nonspecific low back pain: Limited role of imaging'.)
Nevertheless, because numerous trials focus specifically on the presence of Modic type 1 and 2 changes to determine candidacy for interventional treatment, we present similar patient selection criteria. Some clinical trials have shown modest benefit from radiofrequency ablation (RFA) in patients with such findings [41,42]. (See 'Radiofrequency ablation' below.)
●Facet joint arthropathy – Facet arthropathy is characterized by joint space narrowing, osteophytes, sclerosis, and inflammatory changes [43]. However, because the correlation between imaging findings and response to facet blocks is weak, we use a combination of imaging and response to interventional diagnostic procedures to determine candidacy for further interventional treatment. (See 'Interventional diagnostic procedures' below and 'Radiofrequency ablation' below.)
●Isolated muscle pathology – Patients with MRI or other evidence of muscle atrophy in the absence of other features may be candidates for peripheral nerve and peripheral nerve field stimulation, as below. (See 'Other interventional treatments' below.)
MRI is also useful in identifying features suggestive of other back pain etiologies, including lumbar disc herniation or osteophytes causing nerve root compression. Details regarding the treatment of lumbosacral radiculopathy are discussed separately (see "Acute lumbosacral radiculopathy: Treatment and prognosis"). If sacroiliac joint edema or sacroiliac sclerosis is identified, we evaluate for spondyloarthropathy as these findings may be present in both degenerative sacroiliac arthropathy and inflammatory sacroiliitis [44]. (See "Diagnosis and differential diagnosis of axial spondyloarthritis (ankylosing spondylitis and nonradiographic axial spondyloarthritis) in adults".)
Interventional diagnostic procedures
Candidacy — Though the evidence to support their use is mixed, interventional diagnostic procedures are used to identify the most likely pain generator in patients with chronic, nonradicular low back pain, including sacroiliac joint pain [45,46] and facet or zygapophysial joint pain [47-50]. Provocative discography or analgesic discography (aka, "discoblock") may be used to confirm discogenic back pain [51-58].
Selective nerve root blocks are also used to confirm lumbosacral radiculopathy in patients with chronic radicular pain. (See 'Pretreatment evaluation' above.)
Numerous variables complicate the interpretation of published evidence supporting the use of these procedures as a diagnostic strategy. As examples:
●There are no reliable gold standards to confirm the source of pain, and high false-positive rates have been reported with interventional diagnostic procedures [51-53,59,60]. Attempts to reduce high false positives, including requiring a positive response to at least two blocks (eg, "double block") may increase false negatives.
●With any specific procedure, needle placement may vary, depending on the use of anatomical landmarks or radiographic guidance. Injectate volume may also vary and may be administered at volumes exceeding the joint capacity, which undermines the specificity of the intervention.
●Few studies have evaluated whether the use of diagnostic blocks for patient selection improves outcomes when compared with noninvasive diagnostic methods (eg, imaging findings, history, and physical examination) [61]. In one randomized trial of 151 patients with suspected lumbar facetogenic pain, the use of clinical diagnostic methods to identify patients for radiofrequency denervation was associated with greater treatment success than positive response to either single or double diagnostic blocks. However, the overall success rate was lower due to a higher rate of false-negative blocks and excluding placebo responders, and the costs were higher [62].
Procedure — Interventional diagnostic procedures can help confirm the source of pain and serve as predictive tests in anticipation of RFA or surgery. Specific interventional diagnostic procedures are selected based on the interventional specialist's suspicion of etiology. Depending on the procedure, contrast dye, corticosteroids, or local anesthetic agents are used. Most diagnostic procedures are outpatient procedures that are performed with local anesthesia and fluoroscopic or ultrasound guidance. Patients are typically discharged the same day after a postprocedure observation period.
The use of sedation varies among institutions, ranging from no sedation to deep sedation with monitored anesthesia care. For procedures that depend on assessing pain relief after an injection (ie, facet joint nerve [medial branch] blocks and sacroiliac joint injections), we discourage the use of sedation as it can increase the false-positive rate of a diagnostic block [63].
Interventional treatments for patients with successful responses to interventional diagnostic procedures are discussed below. (See 'Interventional treatment for nonradicular pain' below.)
Interventional treatment for nonradicular pain
Etiologies that may respond to nonsurgical intervention — Interventional treatment options for patients with nonradicular pain are reserved for patients with chronic low back pain (eg, >12 weeks' duration) and any of the following conditions, as identified by a combination of clinical evaluation, MRI, and/or interventional diagnostic procedure (see 'Pretreatment evaluation' above):
●Vertebrogenic pain
●Discogenic pain
●Facet joint arthropathy
●Sacroiliac arthropathy
●Lumbosacral radiculopathy
●Muscle pathology (eg, muscle atrophy, taut bands, or muscle spasm on examination or imaging)
Studies demonstrating efficacy are reviewed by indication and intervention, as below. (See 'Radiofrequency ablation' below and 'Glucocorticoid injections' below.)
Interventional treatments do not have a role in patients with nonspecific, nonradicular pain without features of the above conditions. The management of this group of patients is discussed below (see 'Patients without indication for interventional treatment' below). For patients with subacute low back pain (eg, duration 4 to 12 weeks), we prioritize noninterventional treatments [64].
Radiofrequency ablation
Mechanism of action — RFA (also referred to as radiofrequency denervation, neurotomy, or lesioning) involves the destruction of nerves using heat generated by continuous radiofrequency current. Ablation of the associated nerve(s) theoretically halts the transmission of associated pain signals, providing pain relief.
Indications and contraindications — We offer RFA for selected patients with chronic, refractory, nonradicular low back pain and suspected vertebrogenic, facet, discogenic, or sacroiliac joint etiology following the interventionalist's clinical evaluation, MRI, and/or interventional diagnostic procedures (see 'MRI' above and 'Interventional diagnostic procedures' above). Rarely, RFA may be used for discogenic pain that is refractory to more conservative treatments.
Selection criteria have been most consistent in trials evaluating radiofrequency denervation of the intraosseous basivertebral nerve for vertebrogenic back pain. In these industry-sponsored trials, candidates have chronic, nonradicular low back pain and evidence of degenerative disc disease on imaging, typically described as Modic or degenerative endplate changes on MRI [41]. For patients with facet joint, discogenic, and chronic sacroiliac joint pain, trials evaluating radiofrequency denervation are limited by variable selection criteria and denervation technique. As a result, we offer RFA for these patients using a shared decision-making approach, incorporating response to diagnostic procedures and a detailed discussion of risks and limited evidence of efficacy.
Evidence of benefit by indication is as follows:
●Vertebrogenic pain – Trials evaluating intraosseous basivertebral nerve ablation in patients with Modic changes have shown some evidence of both short- and long-term benefits. In two randomized trials, intraosseous basivertebral nerve ablation produced improvements in pain and function [65,66]. In one trial of 225 patients with chronic low back pain and Modic type 1 or 2 changes, RFA of the basivertebral nerves produced a 20.5-point decrease in the average Oswestry Disability Index (ODI) score, compared with a 15.2-point decrease in the sham procedure group at three months; this difference was below the prespecified threshold for clinical significance for the study [66]. A subgroup analysis of responders demonstrated that 75.6 percent of patients receiving basivertebral nerve ablation achieved the minimum clinically important difference (MCID) of greater than a 10-point improvement in ODI score, compared with 55.3 percent of patients receiving a sham procedure. In poststudy longitudinal follow-up, 76.4 percent of patients receiving basivertebral nerve ablation showed sustained improvement in the ODI score at two years [67]. In a subsequent trial of 140 patients randomized to either continued standard of care or RFA of the basivertebral nerve, the study was terminated early due to uniformly improved primary and secondary outcome measures in the ablation arm when compared with patients receiving standard of care [65]. Non-industry-sponsored studies are needed to further guide appropriate patient selection and quantify effectiveness.
●Facet joint pain – Small clinical trials evaluating RFA for facet joint pain [68-75] showed no effectiveness or modest, intermediate-term (<12 months) benefit, with considerable variation in selection criteria and intervention technique that limits generalizability. In a 2015 meta-analysis that included 23 randomized trials, among patients with facet joint pain, radiofrequency denervation modestly improved pain in the short term compared with placebo (mean difference -1.47 on a 10-point scale; 95% CI -2.28 to -0.67; three trials; low- to moderate-quality evidence) but did not improve long-term pain or function [76]. Most of the trials included in the meta-analysis were of low or very low quality, with limited duration of follow-up.
●Discogenic pain – RFA of the ramus communicans has shown some evidence of benefit in clinical trials. In a randomized trial of 49 patients with chronic low back pain at one vertebral level, RFA of the ramus communicans nerve resulted in modest improvement in pain and physical function scores compared with lidocaine injection [77]. Larger studies with more detailed patient selection criteria and longer follow-up are needed to determine the optimal candidacy and treatment parameters for this intervention.
●Sacroiliac joint pain – RFA has shown evidence of benefit for sacroiliac joint pain, specifically in patients with positive response to a prognostic sacral lateral branch block [78,79].
In a small randomized trial of cooled RFA for chronic sacroiliac pain, RFA of the primary dorsal rami and sacral lateral branch nerves between L4 and L5 resulted in 50 percent improvement in pain scores when compared with placebo at six-month follow-up [80]. In a multicenter, randomized trial of 210 patients with sacroiliac joint pain diagnosed by positive response to injection, cooled RFA resulted in modestly improved pain, disability, and quality-of-life scores when compared with standard medical treatment with pharmacotherapy, injections, and integrative therapies through 12-month follow-up [79,81].
In a subsequent series of parallel randomized trials in patients with facet joint disease; sacroiliac joint disease; or a combination of facet joint, sacroiliac joint, or intervertebral disc disease, the addition of radiofrequency denervation did not produce clinically meaningful improvement when compared with a standardized exercise program alone, though the sacroiliac and combination joint disease groups did have modest pain improvement with interventional treatment [75].
RFA is contraindicated in patients with infection at the injection site or systemic infection. While antiplatelet agents and anticoagulation are not absolute contraindications, these medications are occasionally modified or held several days before the procedure.
Procedure
●Initial procedure – RFA is performed by the insertion of an electrode near or on top of the target nerve. The position is confirmed by electrical stimulation or fluoroscopy. Radiofrequency current is then applied to heat and coagulate adjacent tissues, including the target nerve.
With cooled RFA, cool water is circulated through the tip of the probe to create a slightly larger lesion than with conventional RFA, conferring additional theoretical benefit for conditions with variations in the number and location of targeted nerves (eg, sacroiliac joint pain) [82,83].
Both conventional and cooled RFA are performed with local anesthesia, with or without sedation. As an outpatient/ambulatory procedure, the patient can usually return home after a brief period of postprocedure evaluation.
●Repeat ablation – While RFA destroys the target nerves, they may grow back over time, requiring repeat treatment [84]. Guidelines recommend repeating the procedure no more than two times per year. In several studies, repeat medial branch RFA was successful in 60 to 90 percent of patients for whom an initial medial branch RFA improved pain symptoms [85-87].
Adverse effects — Adverse events from RFA reported in clinical trials include increased pain, transient lower limb numbness and paraesthesias (possibly due to neuritis; 5 to 10 percent), new-onset radiculopathy (<1 percent), and superficial burns (<1 percent) [76]. Compression fractures have been reported with basivertebral nerve ablation [63,88,89].
Glucocorticoid injections — We use glucocorticoid injections for selected patients with presumed sacroiliac joint pain based on clinical evaluation, imaging studies, and response to diagnostic procedures, if performed. In three small randomized trials that included a total of 54 patients, both intra- and periarticular steroid injections demonstrated short-term benefits for sacroiliac joint pain [90-92]. A multicenter, comparative-effectiveness study demonstrated some intermediate-term benefits of glucocorticoid injections, though these effects did not achieve the MCID [93]. This intervention is usually considered alongside other interventional and surgical options using a shared decision-making approach.
We do not use glucocorticoid injections in the following anatomic locations for chronic, nonradicular pain, due to limited evidence of benefit, no evidence of long-term benefit, and risk of serious harm:
●Intradiscal injection [94-97]
●Facet joint injection [15,98-100]
●Medial branch injection [15,100-102]
●Local or trigger point injections using either local anesthetic, corticosteroid, or both [103-107]
Based on these studies, the United Kingdom National Institute for Health and Care Excellence guidelines specifically recommend against intradiscal and facet joint glucocorticoid injection in the management of chronic low back pain [61,63,108].
Other interventional treatments
●Peripheral nerve/peripheral nerve field stimulation – In patients with nonspecific, nonradicular back pain with isolated MRI or other evidence of muscle atrophy, peripheral nerve/peripheral nerve field stimulation may be used, though evidence of benefit is limited. These procedures involve the implantation of either temporary or permanent electrodes to treat low back pain presumed to be secondary to muscle atrophy or dysfunction. With these interventions, the activation of efferent fibers innervating dysfunctional muscles is hypothesized to restore normal function, leading to reduced mechanical stress and decreased nociceptive pain signaling.
Two narrative reviews suggest a potential benefit of peripheral nerve stimulation; however, the lack of placebo-controlled trials and minimal long-term efficacy and safety data limit further conclusion [109,110]. In the only placebo-controlled trial, achievement of the primary endpoint (≥30 percent pain relief at 120 days without additional analgesics) was similar between both treatment and control groups [111]. In secondary analyses, the treatment group had a greater reduction in pain scores at 120 days compared with control (<1 point on a 10-point scale); this difference was sustained at three-year follow-up for those who achieved benefit. The use of peripheral nerve stimulation for postsurgical or posttraumatic pain is discussed separately. (See "Interventional therapies for chronic pain", section on 'Peripheral nerve stimulation'.)
●Trigger point injections – Patients with low back pain may have evidence of muscle tension or spasm with palpable trigger points [112], for which trigger point injections may be offered, though the evidence of benefit is mixed and limited by study methodology [107,113]. In one systematic review of 14 studies evaluating botulinum toxin, local anesthetic, and corticosteroid injections, though there was some symptom improvement with trigger point injection therapy, the overall evidence of benefit was inconsistent across a wide variety of injectate types, and no specific injectate demonstrated conclusive benefit [113]. In a separate randomized trial of 378 patients with nonspecific low back pain, paraspinous lidocaine injections combined with standard medical therapy and exercise produced improvement in pain and functional scores over three months when compared with standard medical therapy, exercise, and sham injections [107].
PATIENTS WITHOUT INDICATION FOR INTERVENTIONAL TREATMENT —
Patients with chronic, nonspecific low back pain without evidence of focal pathology (eg, no radicular pain, absence of suggestive imaging, and no response to interventional diagnostic procedures) are viewed by some experts to have a variant of nociplastic pain, or regional pain secondary to central sensitization [114]. In such patients, we refer for a multidimensional treatment strategy analogous to recommendations for patients with a centralized pain syndrome, rather than interventional treatment. Patients with central sensitization are less likely to benefit from interventional nonsurgical therapies and are more likely to experience aberrant responses to diagnostic procedures [15,115].
In rare circumstances, interventional treatments may be offered to these patients as an adjunct to a multidimensional treatment strategy. In such situations, we use a shared decision-making approach, emphasizing the fundamental role of exercise and psychologic support in achieving long-term symptom improvement.
The management of chronic nociplastic pain, including nonpharmacologic and pharmacologic treatment, is reviewed separately. (See "Approach to the management of chronic non-cancer pain in adults".)
INTERVENTIONS WE DO NOT RECOMMEND —
The following interventions have low or indeterminate evidence of benefit:
●Intradiscal glucocorticoid injection – (See 'Glucocorticoid injections' above.)
●Facet or medial branch glucocorticoid injection – (See 'Glucocorticoid injections' above.)
●Piriformis syndrome injection – A subset of patients with sciatic symptoms may have pain related to impingement of the sciatic nerve as it passes through or adjacent to the piriformis or subjacent muscle (eg, gemelli or obturator internus). However, because the diagnosis of piriformis syndrome is not reliably reproducible, the role of injections or other treatments for this condition remains unclear (see "Evaluation of low back pain in adults"). In a randomized trial comparing piriformis muscle injections performed with steroids and local anesthetic or local anesthetic alone, no difference was found between groups [116]. In addition to a possible therapeutic benefit in a small subset of patients, piriformis injections may be performed as a diagnostic procedure.
●Botulinum toxin – The available data for botulinum toxin injections in the treatment of chronic back pain are inconsistent; while, in small randomized trials, paravertebral injection of botulinum toxin A produced mild improvements in pain in the short term, this benefit was not sustained beyond 6 to 12 weeks [117,118]. Further data are needed to confirm findings in a larger number of patients over a longer duration and to evaluate the benefits and harms of repeated injections before this treatment can be recommended.
●Prolotherapy – Prolotherapy (aka, "sclerotherapy") involves the repeated injection of irritants into ligaments and tendinous attachments to trigger an inflammatory response. There is insufficient evidence to support the use of prolotherapy for chronic low back pain [61]. In a systematic review that included five trials of prolotherapy compared with local anesthetic or saline injections for chronic low back pain, there was no difference in short- or long-term pain or disability between prolotherapy and the control intervention in three of the trials [119-122]. Results from one trial that demonstrated a short-term benefit with prolotherapy are difficult to interpret because patients also received a number of cointerventions, including forceful manipulation, injection of tender points, and exercise [123].
●Etanercept – Tumor necrosis factor (TNF)-alpha has been implicated in the pathogenesis of radiculopathy and discogenic back pain. There is insufficient evidence to recommend the use of epidural or intradiscal injections of etanercept (TNF-alpha inhibitor) for lumbosacral radiculopathy; published studies have yielded conflicting results.
In one trial comparing epidural injections of etanercept (4 mg), glucocorticoids, or saline in 84 patients with subacute lumbosacral radiculopathy, one-month pain relief was similar in those receiving etanercept and saline and inferior to steroids [124]. In a phase II, placebo-controlled, dose-escalation study in 49 patients with lumbosacral radicular pain, patients in the low-dose (0.5 mg etanercept), but not higher-dose, group had significant relief of pain at six months compared with placebo [125].
Evidence on intradiscal etanercept (a TNF-alpha inhibitor) is similarly mixed. A small pilot study showed that intradiscal injections of etanercept did not improve pain or disability scores for patients with chronic, nonradicular, discogenic low back pain [126]. A subsequent randomized trial performed in 77 patients with discogenic low back pain found that intradiscal etanercept was associated with improved pain scores at eight weeks (mean difference one point on a 0 to 10 scale), with no difference in function [127]. This study did not evaluate longer-term outcomes.
SOCIETY GUIDELINE LINKS —
Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Lower spine disorders" and "Society guideline links: Radiculopathy".)
INFORMATION FOR PATIENTS —
UpToDate offers two types of patient education materials, "The Basics" and "Beyond the Basics." The Basics patient education pieces are written in plain language, at the 5th to 6th grade reading level, and they answer the four or five key questions a patient might have about a given condition. These articles are best for patients who want a general overview and who prefer short, easy-to-read materials. Beyond the Basics patient education pieces are longer, more sophisticated, and more detailed. These articles are written at the 10th to 12th grade reading level and are best for patients who want in-depth information and are comfortable with some medical jargon.
Here are the patient education articles that are relevant to this topic. We encourage you to print or email these topics to your patients. (You can also locate patient education articles on a variety of subjects by searching on "patient info" and the keyword(s) of interest.)
●Basics topic (see "Patient education: Low back pain in adults (The Basics)")
●Beyond the Basics topic (see "Patient education: Low back pain in adults (Beyond the Basics)")
SUMMARY AND RECOMMENDATIONS
●Patient selection for interventional treatment – Patients should undergo a history and physical examination to identify specific etiologies of low back pain and features that warrant additional workup.
When an underlying etiology is not identified, a presumptive diagnosis of "nonspecific low back pain" is made. Initial management is with at least six weeks of conservative therapies prior to consideration for interventional treatment. (See "Evaluation of low back pain in adults" and "Subacute and chronic low back pain: Management".)
For patients with pain that persists beyond six weeks of conservative therapy, we determine candidacy for interventional procedures based on the presence or absence of radicular features. (See 'Determine if the pain is radicular versus nonradicular' above.)
●Patients with radicular pain – Radicular pain originating in the lower back with radiation down the leg in a dermatomal distribution is suggestive of lumbosacral radiculopathy. It may be accompanied by weakness, reflex changes, or dermatomal sensory loss. Pretreatment studies may include MRI (preferred) or CT. Electromyography/nerve conduction study or interventional diagnostic procedures are also used. (See 'Determine if the pain is radicular versus nonradicular' above and 'Pretreatment evaluation' above.)
For patients with persistent, severe, radicular pain due to a herniated disc, we suggest a trial of epidural glucocorticoids (Grade 2C). Randomized trials of epidural glucocorticoid injections in these patients demonstrate modest short-term improvement in pain and disability, though this benefit is not reliably sustained over the long term.
For patients with symptomatic spinal stenosis and persistent radicular pain, we suggest a trial of glucocorticoid injections (Grade 2C). Although the evidence supporting epidural glucocorticoid injections for symptomatic spinal stenosis is less clear, in our clinical experience, some patients with radiculopathy from spinal stenosis may obtain meaningful relief from epidural corticosteroid injection. Therefore, we suggest a trial of glucocorticoid injections for patients with persistent radicular pain who have not found relief with conservative approaches, particularly in those who wish to avoid surgery or who have contraindications to surgery. (See 'Epidural glucocorticoid injections' above.)
Pulsed radiofrequency of the dorsal root ganglia is an alternative interventional treatment option for patients who do not respond to or have contraindications to epidural glucocorticoids. This intervention is usually considered alongside surgical options (eg, discectomy) using a shared decision-making approach. The evidence of benefit is inconsistent and limited by study methodology. (See 'Radiofrequency interventions for refractory symptoms' above.)
●Patients with nonradicular pain – Patients with persistent, nonradicular low back pain may be candidates for interventional treatments based on the suspicion of underlying etiology following clinical evaluation, MRI, and/or interventional diagnostic procedures. (See 'Pretreatment evaluation' above and 'Etiologies that may respond to nonsurgical intervention' above.)
•Radiofrequency ablation (RFA) – RFA involves the destruction of nerves using heat generated by a continuous radiofrequency current. (See 'Radiofrequency ablation' above.)
For patients with refractory vertebrogenic back pain and compatible Modic type 1 or 2 changes, we suggest RFA of the intraosseous basivertebral nerve (Grade 2C). In randomized trials of patients with chronic low back pain and Modic type 1 or type 2 changes, intraosseous basivertebral nerve ablation was well tolerated and produced some short- and long-term improvement in pain and function, though additional trials with larger sample sizes are needed to quantify effectiveness. (See 'Indications and contraindications' above.)
For patients with refractory pain from presumed facet joint, discogenic, or sacroiliac etiology, we also suggest RFA (Grade 2C). However, the evidence is limited by inconsistent trial results and study heterogeneity. Glucocorticoid injections are also a reasonable alternative treatment for patients with sacroiliac pain, as below.
•Glucocorticoid injections – For patients with chronic, severe low back pain of presumed sacroiliac origin, we suggest glucocorticoid injections (Grade 2C). In randomized trials, intra- and periarticular steroid injections produced modest, short-term improvement in pain; however, these studies were small, and larger studies have not replicated these results. This intervention is usually considered alongside other interventional and surgical options using a shared decision-making approach. (See 'Glucocorticoid injections' above.)
•Other interventional treatments – Although evidence supporting their efficacy is limited, peripheral nerve/peripheral nerve field stimulation and trigger point injections are reasonable alternative treatment options for patients with chronic, nonspecific, nonradicular back pain and evidence of muscle pathology. (See 'Other interventional treatments' above.)
●Patients without indication for interventional treatment – Patients with chronic, nonspecific low back pain who do not have evidence of focal pathology (eg, no radicular pain, absence of suggestive imaging, and no response to interventional diagnostic procedures) are viewed by some experts to have a variant of nociplastic pain, or regional pain secondary to central sensitization. In such patients, we pursue a multidimensional treatment strategy, rather than interventional treatment. (See 'Patients without indication for interventional treatment' above.)
