THE ETIOLOGIES AND IMITATORS OF LUMBOSACRAL RADICULOPATHY

The chief complaint of limb pain greater than axial pain typifies radicular symptomatology, but can also be present in other conditions. Sacroiliac joint dysfunction can present with gluteal and proximal lower limb pain.^2–4 Piriformis syndrome, as described by Yeoman in 1928,⁵ represents a constellation of symptoms including limb pain and/or paresthesias with or without gluteal or lumbar pain.^6–8 Intrinsic hip joint pathology, such as osteoarthritis, will present with gluteal, groin, or anteromedial thigh pain.⁹ Strain injuries to the hamstring musculature will cause complaints of soreness and pain in the posterior thigh. Pain in the region of the greater femoral trochanter, buttock, or lateral thigh depicts greater trochanteric pain syndrome, has a 20% prevalence in patients referred to surgical spine specialists,¹⁰ and can mimic symptoms of lumbar radiculopathy.¹¹ Lumbosacral plexopathy or peripheral nerve entrapment, such as meralgia paresthetica, can also approximate radicular symptoms. Various spinal structures such as the intervertebral disc, dura, ligaments, or zygapophyseal joints may refer pain distally. However, the primary complaint in these instances is axial pain sometimes associated with less severe lower limb symptomatology.

Since Mixter and Barr’s seminal paper in 1934,¹² the intervertebral disc has widely been recognized as a cause of L/S radiculopathy. However, several other etiologies of L/S radiculopathy have been annotated. Central or lateral canal stenosis^13,14 can arise due to an array of causes: congenitally short pedicles, ligamentum flavum buckling, facet joint arthropathy, vertebral body hyperostosis, spondylolisthesis, spinal tumors, or spinal cysts. Intraspinal cysts may evolve as a consequence of the degeneration of the zygapophyseal joint,¹⁵ ligamentum flavum,¹⁶ posterior longitudinal ligament,¹⁷ dura mater,¹⁸ or chronic spondylolysis,¹⁹ and can arise from the root sheath itself.²⁰ Diabetic patients are prone to thoracic radiculopathy,²¹ perhaps due to intraneural absorption of sorbitol while confined by a nonexpansile dural sheath. Extraforaminal compression by the L/S ligaments may also result in lumbar radiculopathy.²²

Analyzing the various culprits responsible for L/S radicular symptoms exposes the intricate nature of the pathophysiology of L/S radiculopathy, and the difficulty in accurately diagnosing and successfully treating these painful conditions.

PATHOPHYSIOLOGY OF LUMBOSACRAL RADICULOPATHY

Mixter and Barr’s landmark description in 1934¹² had led many spine care practitioners to suspect intervertebral disc (IVD) herniation to be causative in a variety of lumbar radicular pain syndromes. The implicit premise has been that biomechanical compression of neural elements was the sole etiologic factor leading to the manifestation of signs and symptoms.²³ However, there is evidence that mechanical influence is not the sole etiologic factor.^24–33 There is little correlation between the severity of radiculopathy and the size of the disc herniation.^25,28,29 Resolution of symptoms after conservative treatment has been observed without a concurrent reduction in disc herniation volume.^28,29 Mixter and Ayer, a year after Mixter and Barr’s hallmark paper, demonstrated that radicular pain could occur without significant disc herniation.³⁰ However, it was not conclusive if this ‘radicular pain’ was nerve-root mediated or somatically referred from another spinal structure. It is probable that, in most instances, biomechanical injury is not the singular cause for the expression of lumbar radicular symptoms related to lumbar intervertebral disc herniation.

Early observations by Lindahl and Rexed³² in 1951 established the presence of pathologic changes including inflammatory cells in nerve roots of patients suffering from sciatica. Subsequent animal studies have demonstrated autoimmune and inflammatory reactions to autogenous nucleus pulposus.^34,35 The human intervertebral disc has been shown to be a potent source of phospholipase A₂,³¹ a regulator of the inflammatory cascade which causes perineural inflammation, conduction block, axonal injury,³⁶ and dorsal root demyelination and mechanically induced ectopic discharges in the rat animal model.³⁷ Herniated lumbar intervertebral discs have been observed to spontaneously produce increased amounts of other potentially neurotoxic inflammatory mediators.³⁸ A rapid transport route may exist, bridging the epidural space and intraneural capillaries, providing quick access for this nuclear material to spinal nerve axons.^39,40

In stark contrast to the peripheral nerve, the nerve root lacks a perineurium, which provides tensile strength and a diffusion barrier.^41,42 Consequently, the nerve root possesses less resilience to tension forces and chemical irritants.⁴² Furthermore, the epineurium, which provides mechanical cushion to resist compression, is less abundant or developed in the nerve root.⁴² Within the nerve root itself the fasciculi do not branch to form a plexiform pattern; instead, they run in parallel, loosely held together by connective tissue.^41,42 Hence, the nerve root is not as well suited to withstand either mechanical or chemical insult as compared to a peripheral nerve. Furthermore, once the inflammatory cascade is initiated, the nerve root lymphatic system is poorly equipped to adequately clear the inflammatory mediators.⁴² An inflamed nerve root is thus predisposed to a chronic inflammatory reaction with invasion by fibroblasts with eventual development of intraneural fibrosis.⁴²

Cadaveric studies have discovered a functional tethering of the nerve root to the intervertebral foramen.^42,43 When an intervertebral disc herniates in a posterior or posterolateral fashion, the exiting nerve root is placed under tension and not always compressed.⁴² The ensuing inflammatory response sensitizes the involved nerve root, decreasing its resilience to biomechanical influences. An inflamed nerve will fire repetitively with just minor perturbations; whereas a nonirritated nerve will tolerate more vigorous manipulation without prolonged firing patterns.^41,44 The length to which a nerve root must be stretched for it to incur neurophysiologic dysfunction is believed to be 10–15% of resting length.^45,46 Clinically, nerve root irritability can be appreciated by elevating the involved lower limb with the knee extended, straight leg raising (SLR). Goddard and Reid⁴³ demonstrated stretch without displacement of the nerve root upon raising the affected limb 20–30° to 70°. Since no nerve root motion is occurring, the radicular pain elicited by this maneuver is a consequence of nerve root tension.⁴³ In asymptomatic patients this movement is nonpainful despite the same amount of tension placed on the neural elements. However, intraforaminal disc herniation or herniation in the presence of spinal stenosis may inflict a biomechanical compromise to the nerve root that may have implications for successful treatment.

The ramification of this growing body of evidence supporting a biochemical paradigm is an evolution of minimally invasive therapeutic interventions. Less than 15% of L/S radiculopathies will eventually require surgical correction.⁴⁷ Thus, a large proportion of cases can be successfully managed conservatively.²⁵ Conservative care includes physical therapy, oral antiinflammatory medications, and precision intraspinal injections; yet, as interventional spine care continues to evolve, further minimally invasive interventions are being offered. Directing a successful plan of care hinges on accurately diagnosing the patient’s condition. Diagnosing L/S radiculopathy does not always create a mystery for the spine clinician. However, proficiently maneuvering through the list of differential diagnoses, utilizing evidence-based medicine, and accounting for the pharmacologic effects of injected medications requires an appropriate algorithmic approach. This systematic evaluation relies upon thorough history acquisition and physical examination, the accurate assessment of imaging studies, astute electrodiagnostic examinations, and the appropriate use of diagnostic block injections.

THE CLINICAL APPROACH TO LUMBOSACRAL RADICULOPATHY

Although radicular pain is categorically defined as limb pain greater than axial pain, occasionally, a patient might report paramedian lumbar pain that could be perceived as back pain. However, this region of discomfort may very well represent the sole region exhibiting painful nerve root symptoms. Yet, the spine clinician must recognize the overlap of referral patterns of pain emanating from nerve root, zygapophyseal joint, intraspinal ligament, intervertebral disc, sacroiliac joint, trochanteric bursa, or the piriformis muscle, and then perform a probability analysis of which structure, or segmental level, is likely generating the symptomatology. Lumbosacral radicular symptoms commonly arise in conjunction with focal intervertebral disc pathology¹² or spinal stenosis.¹³ Disc herniations can lead to functional stenosis and a clinical picture of neurogenic claudication. However, symptom onset associated with disc herniation commonly occurs explosively after a short stint of axial pain or symptoms, and are exacerbated by prolonged sitting or Valsalva maneuvers. Neurogenic claudication due to spinal stenosis progresses more insidiously and its symptoms are exacerbated by prolonged standing and walking. Typically, patients will report bilateral or unilateral lower limb paresthesias, weakness, fatigue, or heaviness that is absent at rest and precipitated by walking that eventually persuades them to stop and rest.⁴⁸ The patient’s walking tolerance, the point at which pain forces the patient to stop and rest, is usually twice the distance at which discomfort is first felt.⁴⁸ Assuming a forward stooping posture may dissipate some of the symptoms, and the capability to walk further while traveling uphill, in contrast to downhill, can discriminate between neurogenic and vascular claudication.

No clinical features characteristically identify sacroiliac joint-mediated pain.⁴⁹ However, a history of a direct fall onto the buttock, rear-end motor vehicle accident (with the ipsilateral foot on the brake), and a fall into a hole (with one foot in the hole and the other extended outside) provides a potential mechanism of injury.⁴⁹ Patients suffering from greater trochanteric pain syndrome have difficulty sleeping on the affected side,¹⁰ and may experience worsening symptoms during ambulation or squatting. Piriformis syndrome is most commonly due to overuse or trauma.⁸ Magnetic resonance imaging studies have suggested the presence of effacement of the fatty sciatic foramen in elite athletes diagnosed with piriformis syndrome.⁸ Lumbosacral plexopathy can be due to pelvic trauma, hip arthroplasty, pelvic or gastrointestinal neoplasms, and autoimmune or vascular disorders.¹ Meralgia paresthetica, lateral femoral cutaneous neuropathy, can be due to compression by tight belts, corsets, seatbelts, psoas muscle tumor, or prolonged hip flexion.¹

Physical examination findings can be helpful in formulating a differential diagnosis, and help guide the diagnostic work-up. Yet, the distribution of radicular pain in radiculopathy appears to be the only sensitive, thus useful, sign indicating the segmental level of disc herniation.⁵⁰ In the authors’ experience at the Penn Spine Center, upper buttock radicular pain has proven to be a manifestation of L4 nerve root-mediated pain, midgluteal pain a representation of L5 root pain, and lower gluteal pain an expression of S1 root pain. Limb pain referred from the L4 nerve root may trigger painful symptoms solely in the anterior knee and/or medial ankle, while L5 nerve root pain may masquerade as isolated lateral knee and/or dorsal foot symptomatology (Fig. 82.1). Although lateral thigh and lateral calf pain characterizes L5 nerve root involvement, posterolateral thigh and posterolateral calf pain implicates S1, or L5 and S1 fiber alterations. Meticulously differentiating between lumbar pain and buttock pain is imperative as patients may refer to both synonymously, but to the interventional spine clinician each indicates different potential pain generators. Employing various dural tension maneuvers will help illuminate the pattern of pain referral into the lower limb. Straight leg raising (SLR), reproducing radicular pain between 20–30° and 70°,⁴³ seems to be sensitive, while crossed SLR is more specific for the level of disc herniation.⁵⁰ Reverse SLR with the patient prone, will assess upper lumbar nerve root fibers. Sensory impairments are less reliable and the diagnostic value of historical findings remains unclear.⁵⁰ Vibratory sensation abnormalities in dermatomal distributions may be the most reliable measure of sensory disturbances.⁵¹

Fig. 82.1 Schematic representation of isolated areas of lower lumbar radicular pain. (A) Lateral view. (B) Dorsal view. (C) Ventral view.

Repetitive motor examination maneuvers such as repetitive calf raises, will elicit subtle motor weakness not detected on static manual muscle testing. When assessing strength statically, the examiner should place the muscle at a biomechanical disadvantage to completely reveal a deficit. Typically, L2 myotomal strength is assessed by hip flexion, L3 by knee extension, L4 by ankle dorsiflexion, L5 by great toe extension, and S1 by ankle plantar flexion. Yet, an astute spine clinician must be aware of anomalous innervation patterns in transitional spinal segments. In instances of sacralization of L5, myotomal innervation patterns overlap so that an L5 radiculopathy may present clinically as L5, S1, or both. Conversely, in lumbarization of S1, innervation patterns are more discrete and clinical findings correlate more closely with the expected compromised nerve root.⁵²

DIAGNOSTIC EVALUATION

Treatment failure may reflect an inaccurate diagnosis, ineffective treatment, or misguided intervention. Lumbosacral radicular symptoms that have not responded to conservative care or are incapacitating for the patient warrant further diagnostic studies to clarify the diagnosis and redirect treatment strategies. Accurately diagnosing the etiology of lower limb pain requires astute and thorough history gathering and physical examination, and the appropriate prescription of imaging studies, electrodiagnostic evaluation, and precision diagnostic injections. These diagnostic evaluations are appropriately divided into visual anatomic, neurophysiologic, and functional diagnostic assessments.⁵³ Plain radiography, radionuclide imaging, myelography, computed tomography, and magnetic resonance imaging constitute the visual anatomic studies. Neurophysiologic tests include electromyography (EMG) and nerve conduction studies (EDX). Anesthetizing injections have been termed functional tests because patient participation is essential to identify the abolition of their typical pain complaints (see Chs 16, 17, 18, 19).⁵³ Although these diagnostic tools have been thoroughly discussed, the following will highlight their clinical utility in evaluating L/S radiculopathy.

Functional diagnostic tests

If the electrodiagnostic testing is inconclusive, advanced imaging demonstrates multiple abnormalities or no corroborative abnormality, fluoroscopically guided, diagnostic selective nerve blocks are performed to elucidate the level of pain generation, and to direct invasive therapeutic measures, whether they are interventional or surgical.⁵³

Diagnostic selective nerve root block

The rudimentary conviction supporting the role of a diagnostic block is that anesthetizing a structure will relieve symptoms if that structure is responsible for those symptoms.⁵³ The accuracy of diagnostic selective nerve root blocks (SNRBs) depends upon the selective anesthetization of the targeted nerve root. In achieving nerve root block, neighboring structures must be avoided to prevent a false-positive response.

As already mentioned, structures other than the lumbar nerve root can be responsible for painful lower limb symptoms. Other structures such as the intervertebral disc, posterior longitudinal ligament, dura, zygapophyseal joint, and sacroiliac joint are known to cause lower limb symptoms mimicking radiculopathy. Kellegren⁷⁸ was first to establish pain referral patterns from non-neural spinal structures when he observed lower limb symptoms upon stimulation of the axial and limb muscle, tendon, fascia, and periosteum. Although Kellegren demonstrated a dermatomal referral pattern with stimulation of spinous process periosteum, Kuslich et al.⁷⁹ found that stimulation of the intervertebral disc anulus, longitudinal ligament, and zygapo-physeal joint provoked a nondermatomal referral pattern.

Subtle similarity between nerve root-mediated pain and the somatic referral from a non-neural structure has significant clinical ramifications. The astute spine interventionist will be able to discern between the two scenarios to enact a successful treatment regimen. The literature does not provide evidence of pathognomonic historical features, or physical examination or radiologic findings to identify the intervertebral disc,⁸⁰ facet joint, or sacroiliac joint^48,81–84 as the etiology of pain referred to the lower limb. Rather, there is support^48,82–85 for the use of anesthetic block of the suspected joint as a valid diagnostic tool. If a diagnostic SNRB is negative in a patient with atypical radicular symptoms or in a patient without a corroborative lesion on visual anatomic studies, an alternate sclerotomal referral source ought to be pursued.⁵³

Nerve root involvement itself may refer pain in an atypical, nondermatomal pattern. Slipman et al.⁸⁶ demonstrated that mechanical stimulation of a cervical nerve root frequently produced symptoms outside of the classic dermatome, in a dynatomal pattern, for that nerve root. Similar referral patterns may arise from the lumbar spine. For example, the furcal nerve is a separate nerve root with its own dorsal root ganglion, and most commonly travels with the L4 nerve root at the L4–5 level. However, it may also accompany the L3 and L5 nerve roots through their respective foramina.⁵³ After exiting the foramen, this anomalous root merges with the lumbosacral plexus, the obturator nerve, or the femoral nerve. Moriishi et al. observed intersegmental anastomoses in the lumbar spine in 20% and 5% of the dorsal and ventral nerve roots, respectively.⁸⁷ However, extradural anomalous nerve roots may have greater clinical significance.⁸⁸ For example, in the instance of a type 2B nerve root anomaly, two roots exit via one foramen, decreasing their resilience to withstand a mass lesion, and obscuring their clinical expression.⁸⁸ Hence, L/S radicular pain may present with dynatomal patterns in a similar fashion as the patients studied by Slipman. Consequently, radicular involvement at one level may manifest as unusual distribution of symptomatology. Nonetheless, a diagnostic SNRB can reliably detect the level of pathology despite an atypical symptom complex.^53,62–68

Utility of diagnostic selective nerve root block

Specificity of a test is the ratio of true negatives divided by the sum of true negatives and false positives, and reflects the probability of a test screening negative in the absence of disease. The specificity of diagnostic SNRBs relies upon the instillation of an anesthetic agent selectively adjacent to the appropriate nerve root without overflow onto neighboring structures or neural elements. The medial branch of the dorsal primary ramus, which provides afferent sensory information to the central nervous system, travels within a few millimeters of the nerve root exiting its foramen.⁸⁹ The sinuvertebral nerve is a recurrent sensory nerve that innervates the dura and posterior longitudinal ligament traveling back through the foramen after branching from the spinal nerve.⁸⁹ The inadvertent anesthetization of these sensory nerves can result in a false-positive SNRB in the presence of a painful dural sheath, posterior longitudinal ligament, or zygapophyseal joint. Precise and accurate technique is, therefore, indispensable in achieving adequate specificity (Figs 82.2, 82.3).

Fig. 82.2 Fluoroscopic image of a right L5 diagnostic SNRB. Notice the thin and precise outline of the exiting right L5 nerve root. This contrast pattern was obtained by positioning the needle tip inferior and lateral to the midpoint of the L5 pedicle as the exiting nerve obliquely traverses the L5 foramen.

Fig. 82.3 Fluoroscopic image of a right L5 transforaminal epidural steroid injection. In contrast to Fig. 82.2, more epidural spread of contrast was achieved in this injection. A more ventral placement of the needle tip into the neural foramen introduced the contrast medium more superiorly and medially into the epidural space.

The specificity of diagnostic SNRBs ranges 87–100%.^{57,65,67,68,90,91} In 1973, Schutz and colleagues found that 13 of 15 patients (87%) with a positive diagnostic SNRB had a corroborative lesion revealed during surgery.⁹¹ However, this retrospective study did not provide information regarding postsurgical outcomes. A year later, Krempen and Smith⁹⁰ reported the surgical results of 16 failed back surgery patients who had both a positive diagnostic SNRB and a corroborative lesion. All 16 patients had previous laminectomies or laminectomies and fusions, and all experienced good or excellent relief of their lower limb pain. In 1985, Haueisen and coworkers⁵⁷ retrospectively analyzed SNRB, electromyography, and myelography results in 55 (57% failed back surgery patients) patients who had positive diagnostic SNRBs. Surgical confirmation of a corroborative lesion was demonstrated in 94% of patients. Dooley et al., in 1988, retrospectively evaluated the responses to diagnostic SNRB and surgical outcomes in 46 patients (51.6% failed back surgery patients).⁶⁵ Forty-four of 46 patients in whom the diagnostic SNRB initially provoked concordant lower limb pain with subsequent, complete relief after the instillation of local anesthetic, had surgical confirmation of corroborative pathology (one patient did not undergo surgery). Interestingly, 9 of 9 patients (100%) with herniated discs experienced complete relief of limb pain, 14 of 17 (82%) with stenosis described complete symptom resolution, the remaining 19 patients had either arachnoiditis or epidural fibrosis, with 8% of the former and 71% of the latter obtaining postsurgical symptom relief. In 1990, Stanley et al.⁶⁸ conducted a prospective study in which 50 consecutive radiculopathy patients underwent diagnostic SNRB. Nineteen of 20 patients who experienced concordant limb pain during nerve root stimulation and subsequent relief following local anesthetic infiltration underwent surgical intervention. Eighteen (95%) had corroborative lesions identified intraoperatively, of which lateral canal stenosis was the predominant abnormality. The authors did not comment on surgical outcomes. In 1993, van Akkerveeken⁶⁷ recorded positive diagnostic SNRBs after symptom provocation with eventual relief after injection of local anesthetic. Of those patients with complete relief of limb pain, surgery confirmed a lesion in 100%. Van Akkerveeken observed a specificity of 90% when negative diagnostic SNRBs were performed at levels free of radiologic evidence of neurocompressive lesions, and a positive predictive value of 95%. However, the positive predictive value may actually fall anywhere from 70% to 95% due to a number of his patients refusing surgical intervention.

The probability of a true-positive test result in the presence of disease reflects the test’s sensitivity. The sensitivity of diagnostic SNRBs has not been as rigorously studied. The diagnostic utility of a clinical test requires that this test be capable of detecting disease at a rate equal to or higher than other available tests.⁵³ Van Akkerveeken⁶⁷ recorded a sensitivity of 100% for diagnostic SNRBs. Haueisen and colleagues’⁵⁷ retrospective study compared the sensitivity of diagnostic SNRB to myelography and electromyography, and observed 99% sensitivity for diagnostic SNRBs in contrast to a rate of correct identification of 24% and 38% for myelography and electromyography, respectively. However, the article did not disclose the authors’ criteria for a positive (or negative) electrodiagnostic study. Regardless, if the sensory afferent axon fibers are primarily affected, an EMG study may not be able to indicate the segment involved. However, in the case of S1 involvement, H-reflexes will reflect sensory fiber dysfunction in the absence of motor fiber injury.

Since the inception of MRI, this diagnostic modality has proven to be very sensitive in depicting spinal abnormalities.^27,33 Yet, determining the clinical significance of these imaged abnormalities requires a complementary diagnostic test of high specificity. Diagnostic SNRBs have great value in verifying a given MRI abnormality as the pain generator when the patient’s symptoms are in conflict with the MRI findings, or in the presence of more than one MRI abnormality.^66,67 Although most available studies antedate the routine use of MRI, several studies have investigated the role of diagnostic SNRB in complex spine cases.

As early as 1971, Macnab⁶² demonstrated the value of diagnostic SNRBs in the preoperative evaluation of patients with negative imaging studies despite clinical signs of nerve root irritation. In 1988, Kikuchi and Hasue⁹² correlated the results of myelography and SNRB with surgical findings and outcomes. Eighty-six out of 119 treated patients underwent surgery and experienced complete resolution of their limb symptoms. Although myelography illustrated the involved segment, contrast patterns during SNRB more accurately localized the abnormal area. Thus, SNRB was capable of identifying a single level of symptomatic root compression in light of imaging and symptoms suggesting multilevel pathology. Decompressive surgical intervention was then limited to the level involved. White⁹³ briefly discussed the utility of preoperative diagnostic SNRB in his 1983 article on diagnostic injections. He supported presurgical diagnostic SNRB as a more precise technique than caudal or interlaminar approaches, and useful in patients with equivocal visual anatomic findings. Herron,⁶⁶ in his 1989 article, retrospectively examined the utility of diagnostic SNRB in selecting patients for surgery after other diagnostic testing – myelography, CT, EMG, and MRI – were found to be equivocal. He found diagnostic SNRBs to be useful in identifying previously undetected disc herniations, the symptomatic level in multilevel disc herniations, the primary pain generator in ‘spine hip syndrome,’ root irritation in spondylolisthesis, the symptomatic level in multilevel stenosis, and the symptomatic root in patients with postoperative fibrosis. However, Herron did not use contrast media to confirm needle location; rather, he relied on concordant symptom provocation during needle placement and 75% postinjection reduction of the patients typical pain. This technique has more important ramifications regarding specificity as compared to sensitivity, however. Stanley et al.⁶⁸ proposed that diagnostic SNRB is also useful in identifying the symptomatic level in patients with anomalous spinal segmentation.

When performed by well-trained interventional spine physicians, major complications are extremely rare. Major complications related to lumbar SNRB including respiratory and cardiac arrest, seizure, infection, nerve root trauma with permanent injury, dural puncture, perforation of viscous and/or vascular structures, broken needles, and even death have rarely been reported. The potential benefit to the patient must be weighed against these known risks. The incidence of both major and minor complications has been extensively studied. Two large studies^94,95 found the incidence of major complications, related to transforaminal epidural steroid injections, to be zero, and the incidence of minor complications (transient nonpositional headache, increased lumbar or limb pain, or vasovagal reaction) was 9.6% per injection.⁹⁵

Diagnostic SNRBs play an integral role in evaluating patients who present with painful L/S radicular symptoms. In trained hands, these minimally invasive procedures are safe, and provide valuable information that may not be available from other diagnostic tests. The evidence supporting this conviction, however, is not definitive. All the above-mentioned studies were retrospective, except for Stanley’s, which was prospective, but did not report outcomes. The authors’ experience at the Penn Spine Center, however, has been that diagnostic SNRBs, when performed meticulously, are safe, reliable, and useful to determine which specific nerve root is, or is not, involved.

Extraspinal axial diagnostic injections

Minimally invasive diagnostic injections can also be utilized to evaluate other etiologies of lower limb pain in the differential diagnosis for lumbar radicular pain. Lower limb pain referred from the sacroiliac joint can be confirmed by diagnostic intra-articular anesthetic injections.^84,96 Schwarzer et al. diagnosed sacroiliac joint (SIJ) pain with single intra-articular injections of anesthetic and found an incidence ranging from 13% to 30% in low back pain sufferers.⁹⁶ Maigne et al. observed a 35% prevalence rate in 54 subjects of SIJ pain employing a single block of short-acting anesthetic.⁸⁴ However, only 10 (18.5% of the cohort) of these 19 responders had a positive response to a longer-acting anesthetic. Thus, SIJ-mediated pain must be part of the differential diagnosis in a patient complaining of gluteal and lower limb pain with negative spinal imaging and electrodiagnostic testing. More importantly, a false-positive response of up to 47.4% may dilute the sensitivity of single diagnostic SIJ blocks. Therefore, a double block paradigm may be necessary to produce accurate information to diagnose SIJ-mediated pain. However, further prospective studies are required to uphold this presumption.

The diagnosis of piriformis syndrome is one of exclusion. Only after negative spinal imaging and electrodiagnostic evaluation for radiculopathy, reproduction of concordant limb symptoms with percutaneous, intrarectal, or intrapelvic pressure should the spine clinician suspect piriformis syndrome.⁷ The modified H-reflex, as described by Fishman and Zybert⁶ may not be a reliable electrodiagnostic tool to evaluate for piriformis syndrome. Slipman et al.⁷ calculated a positive predictive value of 33% in a small study with stringent inclusion criteria. Symptom relief with injection of the piriformis muscle is considered to be diagnostic of this disorder,^7,98,98 and these authors use a symptom reduction of 80% to define a positive injection.

Intrinsic hip pathology can mimic upper lumbar radicular pain or somatically referred pain from the lumbar intervertebral discs.⁹⁹ Faraj et al. retrospectively evaluated responses to diagnostic intra-articular hip injections in 47 patients with radiographically confirmed hip osteoarthritis.⁹ Twenty-four patients had a positive diagnostic block response and subsequently experienced pain relief after total hip replacement. Three out of the 23 nonresponders eventually gained pain relief after total hip arthroplasty 2 years after diagnostic injection. The authors concluded that intra-articular hip injections of local anesthetic are valuable diagnostic aids, and calculated a sensitivity of 88% and a specificity of 100%.

Diagnostic intra-articular hip injections can be employed reliably to differentiate intrinsic hip joint pain from spinal pain. Intra-articular diagnostic SIJ, and intramuscular diagnostic piriformis injections are probably useful in evaluating these structures as the source of referred lower limb pain. Yet, these functional diagnostic tests must be appropriately pursued to confirm one’s clinical suspicion and not relied upon solely to determine the source of a patient’s pain.

AN ALGORITHMIC APPROACH

The spine clinician’s methodology to accurately diagnose and treat a patient presenting with L/S radiculopathy should rely upon evidence-based medicine. At the Penn Spine Center, the authors’ algorithmic approach to L/S radicular symptoms utilizes available evidence to properly diagnose and direct an appropriate plan of care to expeditiously return the patient to his or her prior level of function.

If clinical and advanced imaging findings do not correlate, or if the clinical impression is not specific for monosegmental involvement, an electrodiagnostic evaluation is performed. Electrodiagnostic studies are completed as a functional correlate to the anatomic studies, and to aid in prognosis, as previously outlined. Nerve conduction studies can confirm the presence of an underlying peripheral neuropathy, as well as a peripheral entrapment neuropathy including meralgica paresthetica, and peroneal neuropathy at the fibular head, L/S plexopathy, S1 radiculopathy (H-reflex), and piriformis syndrome (modified H-reflex). The CMAP amplitude of the weakest muscle will reflect the remaining viable axons, help differentiate neurapraxia from axonotmesis, and aid in prognosis. Electromyography assesses the integrity of the motor unit fibers and will help differentiate whether plexopathy, alpha motor neuron disease, diabetic amyotrophy, peripheral neuropathy, entrapment neuropathy, and radiculopathy-provided motor fiber axons are involved. Subsequent nerve conduction and electromyography studies can be performed to semiquantitatively assess the progression or improvement of the neurologic condition. If the electrodiagnostic evaluation conclusively demonstrates L/S radiculopathy at the segmental level of structural disease indicated by advanced imaging, therapeutic interventions are prescribed. Alternate diagnoses are similarly treated if diagnosed by electrodiagnostic examination.

If the electrodiagnostic testing is negative or inconclusive, diagnostic blocks are employed to confirm the clinical impression, and to maximize therapeutic interventions. A diagnostic SNRB is performed at the level of suspicion substantiated by the history and physical examination. However, if advanced imaging and electrodiagnostics are negative, and history and physical examination are supportive, diagnostic blocks of other structures are performed such as the sacroiliac joint, piriformis muscle, hip joint, or trochanteric bursa. In performing a diagnostic SNRB at the Penn Spine Center, the authors use 2% preservative-free xylocaine and define a positive block as an 80% reduction of the preinjection visual analog scale (VAS) rating. A reduction of less than 80% may require the patient to undergo a second diagnostic SNRB with 4% xylocaine if radiculopathy is a firm clinical impression. Since 4% xylocaine does not come preservative free, the authors are particularly cautious about the pattern of contrast spread. If there is any suspicion the injected agents could reach the epidural space, the needle is repositioned until a safe flow is assured. Relying on precision spinal injections to verify the pain generator is appropriate as therapeutic injections probably have distant systemic effects. Thus, if a selective nerve root glucocorticoid injection is moderately successful for a short period of time, the question still remains unanswered: was this effect due to systemic absorption after inaccurate deposition, or did the pathology simply not respond to the appropriately placed anti-inflammatory agent? This decision-making process is more crucial when evaluating the postsurgical patient suffering from recurrent symptomatology. Utilizing this algorithmic approach (Fig. 82.4) at the Penn Spine Center, the authors have been able to accurately diagnose 94.4% of patients presenting with recurrent lumbar and/or radicular symptoms after lumbar surgery.¹⁰⁰

Fig. 82.4 A set of diagnostic algorithms to guide the clinical approach to the patient with a chief complaint of lower limb pain. (A) Algorithm for radiculopathy due to disc herniation. (B,C) Algorithms for central spinal canal stenosis, congenital and acquired, related root injury. (D) Algorithm for radiculopathy due to lateral canal, foraminal, and extraforaminal stenosis. (E) Algorithm for postoperative radiculopathy.

Patients presenting with signs or symptoms of L/S radiculopathy undergo a thorough patient–physician interview and physical examination. Important historical cues include the onset and duration of symptomatology, as well as exacerbating and alleviating factors. Acute onset of explosive limb pain is suggestive of an acute intervertebral disc herniation, and often may be preceded by a short period of midline, axial lumbar pain. In contrast, gradual onset of limb pain may be secondary to nerve root irritation due to degenerative disc disease, intraspinal cysts, facet joint arthopathy/synovitis, tumor, infection, or, in some cases, acute disc herniation. A shorter duration of symptoms, less than 4–6 weeks, may suggest a relatively acute to subacute inflammatory process that may be amenable to antiinflammatory medications. More chronic symptoms, longer than 3–6 months, may still respond to such measures, or may reflect those cases that were destined to fail conservative care, thus requiring surgical correction. Assessing the exact location of limb pain can be difficult as it requires effective communication, and the patient may not be able to articulate this precise information. Evoking referred pain patterns during dural tension maneuvers can be an effective measure to verify a patient’s exact complaint. Eliciting the exacerbating and alleviating factors can also be difficult but must be obtained to better validate advanced imaging findings when multiple abnormalities are discovered, and to better treat the patient’s condition. Activities that increase subarachnoid pressure such as sitting, squatting, coughing, sneezing, Valsalva, lifting, and bending forward will precipitate symptoms due to a disc herniation. In such instances, symptoms can be eased by lying supine or in the lateral decubitus position, changing positions, and ambulation. In contrast, stenosis-related L/S radiculopathy, neurogenic claudication, is triggered by standing, walking (especially downhill), lying supine, and can be reduced by leaning forward or sitting, but prolonged sitting can separately aggravate the symptoms. Vascular claudication should be differentiated from stenosis-related symptoms, as treatment obviously varies. Lower limb pain precipitated by exercise and not relieved by forward flexion, sometimes associated with resting foot pain, warrants a vascular work-up. Red flags such as nocturnal pain worse with recumbency, recent weight loss, history of prior cancer, and immunosuppression must alert the spine care professional to the possibility of tumor or infection. Typifying the behavior of the patient’s L/S radicular symptoms helps categorize a diagnosis allowing the extraction of more useful information from imaging studies, appropriately direct electrodiagnostic studies, and more effective treatment of the patient’s ailment.

Examination of the patient serves several purposes: to exclude long-tract signs, assess weakness patterns, evaluate propensity for consequential injury due to weakness, record provocative maneuvers, gauge legitimacy of complaints, and focus the differential diagnoses. Upper motor neuron signs indicate central nervous system involvement, thus redirecting imaging studies toward the remaining neural axis as deemed by other examination findings. For example, cervical and perhaps intracranial imaging might be indicated in the presence of upper limb motor neuron signs. Myotomal deficits should be discerned from the weakness patterns of plexopathy, neuropathy, myopathy, or myelopathy. Again, such findings will result in different imaging studies. If a myotomal deficit is identified, its functional ramifications must be considered parallel to diagnostic and treatment interventions. A patient with a significant foot drop, for example, may require an ankle–foot orthosis to restore safe and effective gait, and prevent soft tissue injury or a fall. Recording dural tension signs such as the angle at which straight leg raising is positive, provides a baseline for comparison to follow clinical improvement. Subsequent examinations must verify this angle, and that the weakness is stable or improving. If weakness continues to progress, surgery may be indicated. If pain is minimal, and the patient’s chief complaint is weakness and sensory changes, a neurocompressive lesion might be illuminated by advanced imaging. Under such circumstances, decompressive surgical intervention may be necessary to successfully treat the condition.

As has already been detailed, magnetic resonance imaging is very sensitive in detecting spinal abnormalities. Plain radiography can be useful to evaluate osseous alignment. Thus, the authors routinely order both imaging modalities to initiate the work-up of the L/S radiculopathy patient. Lateral flexion–extension views evaluate for segmental instability of levels of degenerative facet joint disease. If gross instability, >4.5 mm of translation at L5–S1 or >4 mm at higher lumber levels; or >15 degrees of sagittal angulation; or >2 degress of rotation is present, surgical care is prescribed. Intervertebral disc height can be assessed prior to performing intradiscal procedures, such as percutaneous discectomy. If MRI displays a corroborative lesion such as HNP, or central or lateral canal stenosis, affecting the clinically suspected nerve root, therapeutic interventions are initiated.

The authors employ a tiered treatment approach when addressing L/S radiculopathy. The initial step involves prescribing physical therapy (PT) with passive modalities for pain relief and graduated exercises that do not increase extremity complaints, oral nonsteroidal antiinflammatory medications (COX-2-specific agents with celebrex 100 mg b.i.d. as drug of choice), avoidance of aggravating activities, and a soft L/S corset if needed. Passive modalities such as cryotherapy, heat, and transcutaneous electrical nerve stimulation help modulate the patient’s pain to allow participation in PT. Lumbosacral stabilization and core conditioning exercises are prescribed in PT to stabilize the L/S spine. William’s flexion-biased L/S stabilization is utilized in cases of stenosis; McKenzie evaluation and treatment is employed in cases of disc herniation. If the patient does not improve over 1–3 weeks, or the patient presents with disruption of the ADLs, insomnia, inability to perform work-related duties due to severe pain, the authors will offer fluoroscopically guided therapeutic injections once the diagnosis has been confirmed. These minimally invasive injection procedures in addition to the aforementioned conservative measures constitute the second tier. The third treatment tier is defined by the use of percutaneous discectomy using the Dekompressor, coblation technology (nucleoplasty) or automated percutaneous discectomy (APLD), in addition to the measures offered in the first two tiers. The authors typically perform percutaneous discectomy in conjunction with a selective nerve root corticosteroid injection to address both the biomechanical and biochemical insults. Percutaneous disc decompression offers a minimally invasive percutaneous treatment, short of surgery, that has been shown to effectively treat L/S radiculopathy.¹⁰¹ In certain cases, however, nucleoplasty¹⁰² may be offered earlier in the treatment algorithm. APLD is preferred when there is a need to extract a larger volume of tissue or in cases of recurrent disc herniation.

If a patient presents with unilateral or bilateral neurogenic claudication after acute onset, and MRI reveals a central focal disc protrusion causing functional stenosis, as can occur in a congenitally small canal due to short pedicles, the authors will incorporate both biomechanical and biochemical treatment pathways. In the authors’ experience, percutaneous disc decompression combined with therapeutic SNRB has been successful in treating such stenosis cases, and if not, the patient is referred to a surgical colleague. In instances of a more gradual onset of claudicating symptoms due to spinal stenosis, its etiology helps guide treatment. Amundsen et al. advocated bracing and physical therapy as initial treatment of stenosis patients presenting with unilateral or bilateral neurogenic claudication.¹⁰³ The authors found that moderate pain would satisfactorily decrease in 50% of the patients after 3 months of conservative care. The other 50% nonresponders and those with severe pain benefited from surgical decompression; however, their treatment algorithm did not encompass therapeutic SNRBs. Botwin et al. documented a 50% reduction in visual analog scale scores in 75% of elderly stenosis patients who underwent therapeutic SNRBs 12 months prior, to treat unilateral neurogenic claudication, in addition to therapy and oral antiinflammatory medications.¹⁰⁴ The Penn Spine Center approach blends the findings of both of these studies. The first tier offers PT, oral antiinflammatory medications, bracing, and activity modification. If symptoms do not abate, therapeutic SNRBs are offered unless there is a superimposed disc protrusion in a tight canal, and a subsequent surgical referral if improvement is unsatisfactory. If, however, a patient complains of lower limb fatigue, cramping, or heaviness with ambulation, or pain not alleviated by sitting, the threshold for a surgical referral is lowered.

Spinal stenosis due to degenerative facet joint cysts may require percutaneous aspiration or puncture. Therapeutic SNRBs can be effective because an inflammatory process is probably present. The authors have observed perineural enhancement on fat-suppression sagittal MRI with gadolinium consistent with perineural venous engorgement and/or inflammation in cases of facet joint cyst-mediated radiculopathy (Fig. 82.5). The third treatment tier in facet joint cyst-related radiculopathy involves attempted aspiration of the cyst via an intra-articular approach through the joint. Involution of the cyst wall can be achieved via suction through a 20-gauge needle with a 10 cc syringe.¹⁰⁵ If aspiration proves unsuccessful, cyst puncture can be attempted, again utilizing an intra-articular approach. As many as 50% of successfully treated facet joint cyst-related radiculopathies may require aspiration or puncture.¹⁰⁶

Fig. 82.5 (A) Sagittal view of T1-weighted fat suppressed images illustrating increased perineural signal around the L5 nerve root due to edema or venous congestion (white arrow). In contrast, the nonenhanced L4 nerve root exits its foramen one segment above (gray arrow). (B) Axial T2-weighted image of the same segment demonstrating the cyst emanating from the right L5–S1 zygapophyseal joint (gray arrow).

In disc herniation-induced radiculopathy, treatment options again follow the tiered protocol, incorporating minimally invasive procedures, if necessary. Regardless of whether the herniation is a protrusion, extrusion, or sequestration, therapeutic SNRBs are offered if more conservative efforts fail. If therapeutic SNRBs are not curative, nucleoplasty, Dekompressor, or APLD can be employed to treat contained herniations. The authors have used the Dekompressor and APLD for disc extrusions with excellent outcomes, but those results have not been published as of yet. Nucleoplasty ought not be performed in the setting of a disc extrusion. If the patient’s pain improves but his or her weakness persists, as has been witnessed in cases of disc sequestration, surgical decompression has often been necessary. Seventy percent of patients will experience significant reduction in pain at 12 months after undergoing 1–4 therapeutic SNRBs^107,108 A smaller proportion, who have disc sequestrations, will experience a similar outcome.²⁵

CASE REPORT

The following case report exemplifying the Penn Spine Center approach to a complicated L/S radiculopathy case.

A 59-year-old female presented with a 6-month history of right paramidline lumbosacral pain referring to her superior buttock and posterior thigh to her midcalf. Her symptoms started gradually, were sharp, burning, and achy in nature, and her VAS rating was a 5 out of 10. Exacerbating factors included sitting, bending forward, driving, sneezing, and coughing. Ambulation consistently alleviated her symptoms. Her pain had not lessened despite physical therapy and Bextra 20 mg b.i.d.

Physical examination revealed a mildly obese female ambulating with a nonantalgic gait pattern without demonstration of Trendelenburg gait or truncal list. Lumbosacral range of motion was full but painful in flexion. Mild tenderness to palpation was noted over the right paraspinal region. Straight leg raising to 50° produced right lumbar and gluteal pain. Sustained bilateral hip flexion reproduced her lumbar and gluteal pain. Manual muscle testing did not detect weakness, and muscle stretch reflexes were intact and symmetric. Sensory examination for light touch, pin-prick, and proprioception were intact. Upper motor neuron signs were absent.

Magnetic resonance imaging revealed intra-articular fluid and joint gapping in the right L4–5 zygapophyseal joint (Fig. 82.6). Subarticular stenosis was noted on the right at this level, and a small central focal protrusion was observed at L5–S1. Flexion–extension plain radiography did not reveal gross instability. An electrodiagnostic evaluation demonstrated normal sural sensory and peroneal motor nerve conduction studies. However, the H-reflex latencies were asymmetric with the right latency 1.45 ms longer than the left.

Fig. 82.6 (A) Sagittal T2-weighted view of the right paramidline lumbosacral spine demonstrating increased signal (black arrow) within the right L4–5 Z-joint. (B) The corresponding axial segment displays intra-articular gapping and increased signal (black arrow).

The differential diagnoses included right S1 versus L5 radicular pain, L4–5 versus L5–S1 Z-joint synovitis, and lumbosacral internal disc disruption syndrome with somatic pain referral. Advanced imaging revealed evidence to support each of these conditions. However, the electrodiagnostic findings suggested a right S1 radiculopathy. The authors’ algorithmic approach employed an initial right S1 diagnostic selective nerve root injection. An L5 diagnostic block followed by diagnostic intra-articular injections of the Z-joints would ensue. However, the S1 diagnostic root injection was positive, and she subsequently underwent two therapeutic right S1 selective nerve injections with complete resolution of her pain.

Purposely withheld from this case report was the history of an ineffective intra-articular steroid injection into the right L4–5 Z-joint months prior to the evaluation of this patient. Increasing the volume within this joint would not be expected to relieve this patient’s symptoms. Although intra-articular injection of antiinflammatory medication might be expected to alleviate the symptoms, the S1 nerve root was the inflamed structure both clinically and electrodiagnostically. Hence, the most appropriate target for treatment was the S1 nerve root, rather than the Z-joint or intervertebral disc.

CONCLUSION

Lumbosacral radiculopathy is most commonly managed successfully by conservative treatment measures.^{25,47,56,67,68} Among these tools are therapeutic SNRBs or transforaminal epidural steroid injections. The mechanism of action of these interventions is not completely understood but it appears to involve both local (W. D. Chow, personal communication, San Francisco, 2003) and systemic effects.¹⁰⁹ Thus, clinical improvement may occur after the instillation of steroids at the incorrect nerve root level or inaccurate site of pathology due to a non-specific systemic steroid effect. Maximizing the therapeutic benefits of axial steroid injections requires accurate placement of these medications. Therefore, appropriate diagnosis is requisite, and warrants the judicious use of visual anatomic, electrophysiologic, and functional diagnostic testing guided by a sound algorithm.