Limb pain, diminished sensation (numbness), altered sensory perception (paresthesias and dysesthesias) and impaired function due to weakness are exceedingly common complaints in the practice of medicine. Many individuals with one or more of these symptoms, particularly pain in isolation, have musculoskeletal problems. Some of these individuals may have sensory symptoms as well. Although this suggests nerve involvement, it is not uncommon to be unable to find objective evidence of nerve pathology particularly if the sensory symptoms are intermittent and vague in their anatomic distribution. By the same token, many of these patients have the perception of weakness that may result from limitations imposed by pain. Not uncommonly, however, patients with complaints of limb pain, sensory symptoms, and altered function will have focal nerve injuries affecting the nerve roots, lumbosacral plexus, or individual peripheral nerves, the subject matter of this chapter.

The purpose of this chapter is to provide a conceptual framework by which to evaluate and manage patients with focal lower limb complaints. The specific goals are to provide strategies to accurately diagnose and then manage focal nerve injuries. This begins by distinguishing them from the musculoskeletal causes of monomelic symptoms described in Chapter 36. Subsequently, as with all neurologic problem-solving exercises, localization is attempted to nerve roots, plexus or one or more individual nerves. As etiologies of nerve injury vary with anatomic locus, the benefit of localization is to limit differential diagnostic considerations, facilitate etiologic diagnosis and provide optimal management. Consideration of chronologic course and risk factors will aid in differential diagnosis.

The format of this chapter will parallel that of the preceding chapter on analogous disorders of the upper extremities to which the reader is referred regarding relevant anatomy, pathophysiology, and electrodiagnostic (EDX) evaluation. To avoid redundancy, these subjects will only be addressed when there are relevant differences between the upper and lower extremities. A detailed review of the clinical features, etiologies, evaluation, and management of individual focal neuropathies of the lower extremities will be provided. As in other chapters in this book, descriptions will rest on a foundation of published data but will be expanded upon by the personal experiences of the authors.

LUMBOSACRAL NERVE ROOTS

There are a few, clinically relevant differences in anatomy and nomenclature between the upper and lower limbs that require repetition. The organization of nerve roots is in many ways identical to that in cervical spine. One notable exception is that dorsal root ganglia may reside in an intraspinal location within the lumbosacral spine. In some cases, this results in mechanical nerve root compression distal rather than proximal to the dorsal root ganglion, producing a potentially confusing pattern of EDX findings to those unfamiliar with this anatomical variant.¹

In the lumbosacral cord the nerve roots have a more oblique, descending trajectory than their cervical counterparts, due to the differing length of the spinal cord and vertebral column (Fig. 24-1). The nerve roots need to descend for a considerable distance from the conus medullaris through the spinal canal before they exit the spinal canal from their designated foramen. Understandably, the root will exit the foramen from the most rostral position within the foramen possible, immediately beneath the pedicle of the vertebral body with the same numerical designation. As this is typically above the plane in which disc material extrudes, or spondylotic bars are most likely to develop, the tendency is to compress the next nerve root which has not yet exited the spinal column and the one corresponding to the lower of the two vertebrae constituting that particular foramen.

In consideration of these influences, disc material extruding from the C5–6 intervertebral disc would preferentially come in contact with a nerve root lying directly above it (C6) in the absence of a pre- or postfixed plexus as mentioned in Chapter 23.² Similarly, in the lumbosacral spine, pathology of the L4–5 disc is most likely to impinge upon the L5 root, not where it exits the foramen but as it traverses the L4–5 disc on its way to the more caudal foramen formed by the L5 and S1 vertebral bodies (Fig. 24-1). There is one additional consideration. In the cervical spine the C5–6 disc pathology preferentially affects the C6 root as their courses parallel each other. With a lumbosacral disc herniation however, as the descending nerve roots traverse the disc perpendicularly, the nerve root that is preferentially compressed is related to how far medial or lateral the disc material protrudes from the rent in the annulus fibrosis. For example, disc herniation at L4–5 near the midline preferentially compresses a more medially positioned S1 or other sacral nerve roots. Alternatively, the more typical posterior-lateral disc herniation that occurs lateral to the posterior longitudinal ligament may preferentially affect the L5 root. A far lateral herniation may compress the laterally placed L4 root against its pedicle or overlying lamina (Fig. 24-1).

It is important to recognize two other potential variations from typical compressive radiculopathy. Segmental patterns of injury do not necessarily originate from pathology of nerve roots but can originate from injury to analogous segments of the spinal cord, particularly if the pathology is affecting the anterior horn but sparing the centrifugally placed descending motor and ascending sensory tracts. These segmental deficits may also result from compressive cord injury that may be at a level more rostral than the clinical deficits. Hypothetically, this results from an ischemic mechanism similar to what has been proposed in neoplastic spinal cord compression. Lower motor neuron deficits resulting from presumed ischemic anterior horn cell injury has been described in both cervical spondylotic myelopathy and from dural arteriovenous injury.^3,4

It is also true that radiculopathy can be obscured by myelopathy, particularly in the cervical and thoracic regions. A spondylotic bar encroaching on the spinal canal in the neck is more likely to be manifest with tract rather than segmental signs and symptom whereas this same bar in the lumbosacral spine can have radiculopathic but not myelopathic manifestations.

Musculoskeletal conditions are estimated to underlie 70% or more of back pain cases.⁵ Understanding back and radicular pain requires a basic understanding of the lumbosacral spine anatomy (Fig. 24-2).⁶ The articular surfaces that contribute to both the mobility and stability of the spine include the intervertebral discs and two pairs of synovial joints that together form the articular connections between contiguous vertebrae. These synovial joints include the zygapophyseal or facet joints that are formed by extensions of two contiguous vertebral laminae and constitute the roof of the neural foramina. These latter joints are innervated by branches of the posterior ramus of the spinal nerve. Diseases affecting these structures are one of many potential sources of nonradiating back pain. The second synovial joint system, the uncovertebral joints of Luschka, arises from the posterior-lateral surfaces of the vertebral bodies. These joints along with the disc itself constitute the floor or ventral boundary of the neural foramina through which nerve roots exit. The rostral and caudal boundaries of the neural foramen are formed by the pedicles of the vertebrae immediately above and below the foramen in question.

There are two major ligamentous structures within the spinal canal: the posterior longitudinal ligament and the ligamentum flavum. Both are longitudinally oriented, the former running along the anterior aspect of the central canal just posterior to the vertebral bodies and disc spaces. The latter runs along the posterior aspect of the spinal canal just underneath the spinous processes. The posterior longitudinal ligament is half the width of its cervical counterpart in the lumbar spinal canal. This may add to an increased risk of paramedian disc herniation with potential consequence related to bowel and bladder control. The posterior longitudinal ligament is innervated by the sinuvertebral (recurrent meningeal or recurrent nerves of Luschka) nerves that arise from the rami communicantes outside the neural foramina. These travel posteriorly to innervate the dura, annulus fibrosis, the walls of intraspinal blood vessels as well as the posterior longitudinal ligaments (Fig. 24-2). The ligamentum flavum contains few nociceptive fibers. Its major clinical significance may be to contribute to canal stenosis by its tendency to hypertrophy as part of the spondylotic process. The diameter of the central canal averages 18 mm in most normal adults with a range of between 15 and 23 mm. As in the cervical canal, it widens by a few millimeters when the patient bends forward.

Although this is a text of neuromuscular disorders, it is appropriate to mention potential sources of back, buttock, thigh, and leg pain. It is safe to say that isolated back pain without radicular pain or neurologic signs or symptoms may occur as the initial symptom of disorders which may eventually have neurologic consequences. It is equally safe to say that it may be difficult to initially distinguish common non-neurologic and often musculoskeletal causes of back pain from less common ones that have or may develop neurologic consequences. In the former category, potential anatomic sources of back pain include many spinal structures, such as the posterior longitudinal and other ligaments, capsules of the facet and sacroiliac joints, vertebral periosteum, dura, the paravertebral musculature and fascia, blood vessels, annulus fibrosus, spinal nerve roots, epidural veins and arterioles, and epidural fibroadipose tissue.⁵ Although it has long been suggested that paraspinal muscle pain originates from muscular spasm promoting constriction of intramuscular blood vessels, the lack of continuous EMG activity in hardened, tender muscles suggests that myoedema rather than continuous muscle activity promotes back stiffness and discomfort. In any event, identifying the anatomic source of back pain in an individual patient is an extremely difficult undertaking. Due to their lack or relative lack of nociceptive nerve endings, neither the nucleus pulposus nor the ligamentum flavum appear to be likely culprits.

Degenerative spine disease, i.e., spondylosis, affects a number of different structures, which may individually or collectively narrow the diameter of the neural foramen or the central canal of the spinal column. As a consequence, nerve root integrity may be compromised in either location by enlargement of normal anatomic structures. Degeneration of the zygapophyseal and uncovertebral joints promotes osteophyte formation and space occupying joint enlargement. Intraspinal ligaments hypertrophy. Degeneration of the intervertebral disc results in bulging of the annular ring and loss of its vertical height reducing intrapedicular distances and contributing to foraminal narrowing. If spondylolysis and resulting spondylolisthesis occurs, that is the shifting of one vertebral body on another in an anterior–posterior direction, both central canal and foraminal cross-sectional area is compromised.

The intervertebral disc consists of a gelatinous center, the nucleus pulposus, and a cartilaginous margin, the annulus fibrosis. As mentioned, the concept of discogenic pain is somewhat nebulous in that there are a paucity of nociceptive pain fibers innervating the outer annulus and none within the nucleus pulposus itself. Although the pain and pathophysiology of nerve root disease are typically attributed to direct compression of the nerve root and the inflammation that accompanies it, it is important to remember that other potentially pain-sensitive structures such as the sinuvertebral nerves traverse the neural foramina as well.^7–9 Although there is a rich anastomotic blood supply to the spinal cord and nerve roots, ischemic injury resulting from radicular vascular compression may represent an alternative mechanism of nerve root injury.

LUMBOSACRAL PLEXUS

There is considerable variation in the anatomy of the lumbosacral plexus (Figs. 24-3 and 24-4). It may have contributions from as many as 11 spinal nerves but is typically composed of 8 (L1–S3). The lumbar plexus is predominantly composed of branches from L1 to L4, with variable contributions from T11, T12, and L5. Typically, the majority of L4 fibers travel with the lumbar plexus, with a much smaller contribution from L4 joining with L5 to form the lumbosacral trunk. In a “prefixed” plexus, the plexus shifts downward so that there is more of an L1 contribution to the lumbar plexus, the femoral and obturator nerves become comprised of L2–3-rather than L3–4 segmental contributions and the majority of L4 fibers end up in the sacral plexus. In the so-called “postfixed” plexus, the plexus is shifted upward so that virtually all of L4 and some of L5 are now confined within the lumbar rather than sacral plexuses.

The lumbar plexus is formed in the retroperitoneum, just inferior to the kidney and just behind the psoas muscle. Its blood supply originates from the internal iliac artery. Ischemic injury may occur from distal aortic or internal iliac arterial occlusion. The major branches of the upper lumbar plexus are the ilioinguinal, genitofemoral, and lateral femoral cutaneous nerve (LFCN) or lateral cutaneous nerve of the thigh. The femoral, obturator, and lumbosacral trunks are the major components of the lower aspect of the lumbar plexus.

The sacral plexus is formed within the concavity of the ventral surface of the sacrum, behind and lateral to the rectum. The L4 and L5 contributions to the sacral plexus and to the sciatic nerve are provided by the lumbosacral trunk, the conduit between the lumbar and the sacral plexuses. The lumbosacral trunk traverses the pelvic brim at the posterior aspect of the pelvis, over the sacral alae, and just lateral to the sacroiliac joints (Fig. 24-3). In this location, it is vulnerable to compressive injury during parturition. The major branches of the sacral plexus are the superior and inferior gluteal nerves, the posterior cutaneous nerve of the thigh, the fibular (formerly peroneal) and tibial divisions of the sciatic nerve, and the pudendal nerve.

As mentioned, the embryologic rotation of the limb results not only in the spiral orientation of the dermatomes and hip ligaments but in relocation of muscles from their original anatomic positions. Muscles that were originally located on the posterior surface of the lower limb are innervated by the posterior branches of the lumbosacral plexus, for example, femoral, fibular (formerly peroneal), superior and inferior gluteal nerves as well as the lateral cutaneous nerve of the thigh. Muscles that were originally in an anterior location are innervated by anterior branches, for example, genitofemoral, obturator, and tibial nerves.

INDIVIDUAL PERIPHERAL NERVES OF THE LOWER EXTREMITIES

Identification of lower extremity mononeuropathies is dependent on knowledge of patterns of muscle weakness, sensory symptoms and reflex loss if relevant. As the pattern of muscle weakness arguably provides the most objective localizing information, knowledge of the muscles that promote the major movements of the thigh, leg, foot, and toes and the nerves that innervate them is invaluable (Tables 24-1 and 24-2). Determination of potential etiology however is enhanced by a detailed understanding of the relationship between the nerves and contiguous anatomic structures. The following paragraphs summarize the relevant lower extremity peripheral nerve anatomy.

The iliohypogastric nerve is primarily an extension of the L1 nerve root with some contribution from T12. It exits on the lateral border of the psoas muscle in proximity to the lower pole of the kidney and traverses the ventral surface of the quadratus lumborum muscle. It exits the abdominal wall superior to the iliac crest. It provides partial innervation to the transverse abdominus and internal oblique muscles of the abdominal wall. There are two cutaneous branches, one overlying the iliac crest in the posterior axillary line and a second innervating a small transverse patch above the pubic symphysis (Fig. 24-5).

The ilioinguinal nerve has a similar L1 segmental origin and anatomic course. Its course is parallel but caudal to the iliohypogastric nerve along the upper border of the iliac crest. Its course is retrocolic along the posterior abdominal wall. The nerve passes through the superficial inguinal ring to supply the skin overlying the inguinal ligament, extending to the regions just above and lateral to the base of the penis and scrotum (or labia). In other words, the area just above the genitals and just below the pubic symphysis (Fig. 24-5). Like the iliohypogastric nerve, the ilioinguinal nerve innervates the transverse abdominus and internal oblique muscles.

The genitofemoral nerve has near equal contributions from the L1 and L2 segments. It penetrates the psoas muscle in the retroperitoneum and descends vertically along its ventral surface. It lies in close proximity to the external iliac artery, ureters, terminal ileum on the right, and sigmoid colon on the left. Like the iliohypogastric nerve, it has two separate sensory branches. The larger of the two, the femoral branch, innervates the anterior, proximal thigh in the midline, just distal to the inguinal ligament. The second, smaller genital branch, supplies a small cutaneous zone on the lateral aspect of the root of the penis and scrotum or corresponding area of the labia. Its cutaneous distribution overlaps with portions of the ilioinguinal and iliohypogastric territories (Figs. 24-3 and 24-5). The only muscular branch is the cremaster muscle which controls the ascent/descent of the testes in order to maintain spermatic temperature homeostasis.

The obturator nerve receives contributions from the second through fourth lumbar segments (Figs. 24-3,24-5,24-6 and 24-7). It emerges from the medial border of the psoas muscle just rostral to the pelvic brim and descends through the pelvis vertically, medial to the course of the femoral nerve, to exit the pelvis through the obturator foramen. It innervates the adductor longus, brevis, and a portion of the adductor magnus muscle, as well as the gracilis, and obturator externus muscles. The major function of these muscles is to adduct the thigh with contributions to thigh flexion and external rotation. Cutaneous sensation is supplied to a small patch on the inner thigh.

The femoral nerve is also an extension of the L2–4 segments (Figs. 24-3 and 24-7). It arises in a retroperitoneal location and passes between the psoas and the iliacus muscles before traveling under the iliacus fascia in the lateral pelvis, where it is potentially vulnerable to an iliacus compartment syndrome. It exits the pelvis below the inguinal ligament and lateral to the femoral artery. From a motor perspective, the femoral nerve innervates the psoas and the iliacus muscles in the pelvis and six muscles in the thigh, including the four components of the quadriceps, the sartorius, and the pectineus muscles. The primary function of the majority of these muscles is to extend the leg at the knee joint. In addition, the iliopsoas, sartorius, pectineus, and the rectus femoris all contribute to hip flexion. The rectus femoris is the only quadriceps muscle that originates from the pelvis and is therefore the only one of the quadriceps capable of contributing to hip flexion. The sartorius is an unusual muscle as it contributes to external rotation at the hip joint, flexion at the knee joint, and hip flexion. The pectineus muscle contributes both to external rotation and adduction of the thigh. From a sensory perspective, the femoral nerve supplies sensation to the anterior surface of the thigh and the medial aspect of the leg through its terminal sensory branch, the saphenous nerve.

The LFCN is an extension of the second and third lumbar nerve roots (Figs. 24-3 and 24-6). It also emerges from the lateral border of the psoas and traverses the lateral pelvis deep to the iliacus fascia. It exits the pelvis at the anterior superior iliac spine, often penetrating the lateral margin of the inguinal ligament. It has no motor function and provides cutaneous innervation to the anterolateral thigh as well as the underlying fascia.

Prior to the actual formation of the sciatic nerve, there are four nerves originating from the upper sacral segments. The pudendal nerve is the more proximate of these, originating from the S2–4 segments. In a slightly more caudal location, the posterior cutaneous nerve of the thigh is formed by two or more S1–3 segments before these segments merge with the lumbosacral trunk to form the sciatic nerve which occurs just lateral and anterior to the sacrum. The last branches departing the sacral plexus prior to the formation of the sciatic nerve are the superior and inferior gluteal nerves. They are comprised of the L4–S1 and L5–S2 segments, respectively. The superior gluteal nerve is typically the only nerve to exit the sciatic notch above the piriformis muscle, the sciatic, inferior gluteal, pudendal and posterior cutaneous nerves of the thigh all typically exiting the sciatic notch caudal to this horizontally oriented muscle. Intramuscular injections are avoided in the inferior, medial quadrant of the buttocks, in order to avoid injury to these nerves which travel deep to this topographical location. The superior gluteal nerve innervates the gluteus medius, gluteus minimus, and tensor fascia lata muscles. Thigh abduction at the hip joint is their major action. All contribute to internal rotation of the thigh as well. The gluteus minimus provides a minor contribution to hip flexion, and the posterior aspect of the gluteus medius contributes partially to external rotation of the thigh. The inferior gluteal nerve innervates the gluteus maximus, which is the primary hip extensor, but provides a minor contribution to external rotation as well. Neither nerve has cutaneous representation.

The sciatic nerve receives contributions from the last two lumbar roots via the lumbosacral trunk and the first three sacral segments (Figs. 24-3 and 24-4). In reality, it is really two nerves that are conjoined, the tibial and the fibular (formerly peroneal) nerves. As many sciatic neuropathies preferentially affect the fibular nerve and may mimic a fibular neuropathy at a more distal location, it may be helpful to conceptualize the sciatic nerve as two separate nerves. The segmental contribution to these two nerves is somewhat different. The peroneal nerve contains few, if any, S3 fibers, whereas there is no meaningful L4 contribution to the tibial nerve in the majority of individuals. The sciatic nerve exits the pelvis through the sciatic notch, typically beneath the piriformis muscle, but at times traversing through or above it. The former provides the anatomic basis for the controversial piriformis syndrome. The sciatic nerve descends lateral to the ischial tuberosity of the pelvis and medial to the greater trochanter of the proximal femur, where it is potentially vulnerable not only to misplaced injections but also to displaced hip fractures or inadvertent injury during arthroplasty.

In the thigh, the sciatic nerve innervates the hamstrings, the short head of the biceps innervated by the lateral trunk or fibular (peroneal) portion of the nerve. The remaining three muscles; the semitendinosus, semimembranosus, and long head of the biceps; are innervated by the medial trunk or tibial division. The lateral two muscles are largely S1 innervated whereas the medial two muscles are predominantly L5. In addition, the adductor magnus may receive partial innervation by the sciatic nerve, providing a potential source of electrodiagnostic confusion for the unwary. A lesion of the sciatic nerve proximal to the knee will produce a pattern of sensory symptoms or sensory loss that includes the entire foot and the distal half of the lateral surface of the leg, sparing the L4/saphenous innervated medial leg. The blood supply to the sciatic nerve originates predominantly from branches of the inferior gluteal artery and popliteal arteries. This creates a watershed at mid-thigh level, which has been proposed as an explanation for both the location and prevalence of sciatic neuropathies in vasculitis.

The posterior cutaneous nerve of the thigh exits the pelvis through the lower sciatic notch, medial to the sciatic nerve and lateral to the pudendal nerve. Like the sciatic nerve, it may travel through the piriformis muscle in some individuals. It travels deep to the gluteus maximus which protects it. At the level of the gluteal crease, cluneal branches exit and ascend to supply the skin of the inferior buttock. There are perineal branches as well, which supply the skin and fascia of the lateral perineum, the proximal medial thigh, and the posterolateral aspect of the scrotum/labia as well as root of penis/clitoris. The terminal branch descends vertically to provide sensory capability to the posterior thigh and often proximal aspect of the posterior calf. The posterior cutaneous nerve of the thigh has no motor function.

In the leg, the common fibular nerve bifurcates below the level of the fibular head into its deep fibular (peroneal) and superficial fibular divisions (Figs. 24-4 and 24-8). The deep fibular nerve innervates the muscles of the anterior compartment: the tibialis anterior (TA), the extensor hallucis, the extensor digitorum longus, and the peroneus tertius, a muscle of electrodiagnostic interest. In the foot, it innervates a solitary muscle: the extensor digitorum brevis (EDB). Collectively, the major function of these muscles is to dorsiflex the foot at the ankle and the toes at the metatarsal–phalangeal joints although the peroneus tertius contributes to ankle eversion as well. The superficial fibular (peroneal) nerve innervates the lateral compartment of the leg, including the peroneus longus and brevis muscles. The major function of these muscles is to evert the foot at the ankle. The deep fibular nerve has a predominantly motor function with a very small cutaneous contribution to the interdigital space between the first and second digits. The superficial fibular nerve innervates the skin of the dorsal surface of the foot and the distal-lateral surface of the leg.

The tibial nerve receives contributions from the L5–S3 nerve roots and is the continuation of the medial cord of the sciatic nerve (Figs. 24-4 and 24-9). It physically separates itself from its fibular (peroneal) counterpart in the distal thigh, passes through the popliteal fossa before passing between the two heads of the gastrocnemius muscle. As previously mentioned, it innervates three of the four hamstring muscles in the thigh. In the leg, it supplies the posterior compartment including the two heads of the gastrocnemius, soleus, tibialis posterior, flexor digitorum longus, and flexor hallucis longus muscles. In the foot, it supplies all intrinsic foot muscles except the EDB. Its primary functions are to flex the leg at the knee, to plantar flex and invert the foot at the ankle, and to flex, abduct, and adduct the toes. The three cutaneous branches of the tibial nerve all branch at the level of the medial malleolus and include the medial and lateral plantar and calcaneal nerves. These provide the cutaneous innervation for the medial sole, lateral sole, and heel surface, respectively.

The sural nerve is formed in the popliteal fossa by anastomotic contributions from the common fibular and tibial nerves. It is derived primarily from the S1 nerve root. It descends in a fairly superficial, posterior position in the calf, moving somewhat laterally as it passes behind the lateral malleolus. It provides cutaneous innervation to the lateral foot.

The pudendal nerve has a convoluted course. It first exits the pelvis through the greater gluteal foramen only to reenter through a narrow aperture and potential site of entrapment between the sacrotuberous and sacrospinous ligaments. It then passes through Alcock’s canal created by the obturator muscle posteriorly and the ischial tuberosity anteriorly before exiting the pelvis for good just below the symphysis pubis.¹⁰ It has three major branches: the inferior rectal or hemorrhoidal, the perineal, and the dorsal nerve of the penis/clitoris (Figs. 24-3,24-4 and 24-5). The inferior rectal nerve innervates the external anal sphincter and supplies sensation to the distal anal canal and perianal skin. The perineal nerve innervates the muscles of the pelvic floor, the external urethral sphincter, and the erectile tissue of the penis. Its cutaneous innervation includes the perineum anterior to the rectum as well as the scrotum and labia. The dorsal nerve of the penis is a purely sensory branch whose cutaneous representation is the skin of the penis and labia.

PATHOPHYSIOLOGY

The pathophysiology of peripheral nerve injury has been described in detail in Chapters 2 and 22. Axon loss with Wallerian degeneration typically results from disorders that infiltrate or infarct nerves and may result when nerves are sufficiently inflamed or mechanically injured by compression or stretch of adequate intensity or duration. Axon loss is often accompanied by pain, often deep, aching and burning in character. It is common, particularly with acute or subacute pathological processes. Muscle weakness and atrophy, sensory loss that affects all modalities, loss of deep tendon reflexes if relevant to the nerve injured, and even dysautonomic manifestations including sweating and vasomotor abnormalities are anticipated. Electrodiagnostically, amplitudes of involved sensory and motor nerves diminish on nerve conductions and fibrillation potentials and eventually a reduced number of enlarged, reconfigured motor unit action potentials (MUAPs) develop.

Many experimental models of nerve compression support the belief that myelin is preferentially damaged in the early stages of external compression or internal entrapment. Electrodiagnostically, this may express itself by any combination of focal and uniform slowing of affected fibers, nonuniform slowing (i.e., temporal dispersion), or conduction block. Demyelinating nerve injuries are, in general, less painful than their axonal counterparts but this is etiologically dependent. Clinically, focal slowing may produce paresthesias but no objective deficits. Differential slowing may impair modalities that are dependent on the synchrony of impulse transmission such as deep tendon reflexes and the perception of vibration. Conduction block causes weakness without atrophy (other than that attributable to disuse) and loss of sensory modalities dependent on large myelinated fibers including vibration and position sense. Needle examination of muscles innervated by nerves affected by conduction block demonstrates reduced recruitment of MUAPs but neither fibrillation potentials nor abnormal motor unit action potential morphology as there is no axon loss or reinnervation. As implied above, individual fibers may be affected by focal slowing, demyelinating conduction block, or axon loss, leading to an mixed axonal demyelinating EDX pattern.

ELECTRODIAGNOSTIC STUDIES

Electrodiagnosis (EDX) has a significant role in determination of the existence, location, pathophysiology, severity, and prognosis of focal lower extremity neuropathies. Detailed description of EDX as a diagnostic tool can be found in Chapter 2 and many of the principles of EDX relevant to focal neuropathies of the upper extremities found in Chapter 23 and in the previous section apply here as well.

In general, monoradiculopathies are characterized by normal sensory nerve action potentials (SNAPs) in relevant dermatomes, normal or reduced compound muscle action potential amplitudes (CMAP) in relevant myotomes depending on the degree and type of injury, and evidence of acute and/or chronic denervation in muscles innervated by a single segment but by more than one peripheral nerve. Denervation is frequently, but not universally, found in analogous paraspinal segments. Presumably, failure to demonstrate paraspinal denervation in radiculopathy reflects sampling error, and demyelinating pathophysiology or in more longstanding cases, successful reinnervation. Practically speaking, monoradiculopathies that can be confirmed electrodiagnostically have at least some component of axon loss as the ability to identify demyelinating lesions in proximal locations is limited by anatomic and other considerations.

For example, the EDX pattern of an L5 radiculopathy would include a normal superficial peroneal SNAP and normal or reduced CMAP amplitude recording from the EDB or TA muscles and evidence of denervation in muscles such as the TA, the flexor digitorum longus, the tensor fascia lata as well as lumbosacral paraspinal muscles. These share L5 segmental innervation but are innervated by four different peripheral nerves. Like the clinical examination however, not all muscles innervated by the L5 segment will be denervated in all cases or denervated to the same degree.¹¹

Polyradiculopathy has a near identical EDX pattern. The major exception is the pattern of denervation on needle EMG, which is commonly bilateral and by definition found in multiple segmental distributions. The major distinction between polyradiculopathy and multifocal neuropathy or plexopathy is the sparing of SNAPs in polyradiculopathy. There may be variables such as age, intraspinal positioning of the dorsal root ganglia, lower extremity edema, or concomitant but unrelated polyneuropathy which may be confounding. In addition, particularly in chronic polyradiculopathies such as spinal stenosis, denervation may take on a pseudo-length–dependent pattern suggesting a polyneuropathy.¹² CMAP amplitudes are more likely to be reduced as the protection offered to individual muscles by multisegmental innervation is less prevalent.

Polyradiculoneuropathy or radiculoplexus neuropathy is a pattern that is arguably more relevant to the lower extremities in view of the predilection for diabetes to affect lumbar nerve roots and contiguous plexus and nerve elements. It is a pattern that may occur with acquired, inflammatory demyelinating neuropathies as well but these are usually easily distinguished both by phenotype and by the characteristic demyelinating features found on nerve conduction studies. In general, polyradiculoneuropathies are electrodiagnostically defined by concomitant paraspinal denervation and abnormal SNAPs.

Plexopathies are typically monomelic but may affect the contralateral limb concomitantly or on a delayed basis, depending on etiology. Both clinically and electrodiagnostically, the pattern is typically one of both motor and sensory involvement involving multiple nerve and nerve root distributions. Relevant SNAP and CMAP amplitudes are reduced. Denervation will be found in proximal as well as distal limb muscles innervated by the same elements of the plexus, for example, both the tensor fascia lata and the TA in a lumbosacral trunk lesion, but should not occur in the representative areas of the lumbosacral paraspinal muscles. Demyelinating features may occur in plexopathies but are again often obscured by the proximal location of the pathology inaccessible to routine nerve conduction studies. Uncommonly focal, acquired demyelinating neuropathies such as multifocal acquired demyelinating sensory and motor neuropathy (MADSAM or Lewis-Sumner syndrome) may initially occur in a pattern that is both clinically and electrodiagnostically suggestive of plexopathy although this is usually more of an issue in the upper extremities.¹³

With axon loss mononeuropathies, reduced SNAP and CMAP amplitudes are expected assuming they are performed late enough to allow completion of Wallerian degeneration. By definition, these abnormalities will be confined to the affected nerve. Mild degrees of axon loss may be more readily detected by comparing the amplitude of the affected to the unaffected side rather than to population norms. This is particularly true for SNAPs. Most electrodiagnosticians consider an amplitude of less than 50% of the unaffected side to be abnormal. Denervation on needle examination would be confined to muscles innervated by the peripheral nerve in question but may be limited by site of injury along the length of the nerve as well as by selective fascicular involvement. As an example, denervation would understandably occur in the TA but not the peroneus longus muscle in an axon-loss deep peroneal neuropathy. The same pattern of denervation however, could be conceivably found in more proximal neuropathy affecting the common peroneal or even sciatic nerve due to selective fascicular involvement. Nerve fibers destined to innervate specific muscles may be sequested to specific fasicles in proximal nerve locations. As a result, partial nerve injury in a proximal location may result in selective fascicular injury resulting in an incomplete pattern of denervation.^14,15

The EDX pattern in predominantly demyelinating mononeuropathies differs considerably from their axonal predominant counterparts. Again, by definition, the pattern of abnormalities would be confined to a singular nerve distribution. Sensory nerve conductions should be normal unless there is an axonal component to the injury or there is conduction block that exists between the stimulation and recording sites. Demyelination will have no effect on the conductive properties of a nerve if the lesion is either proximal or distal (as opposed to within) the tested segment of nerve. For example, with a demyelinating common fibular neuropathy at the fibular head, the superficial fibular SNAP amplitude obtained from a location distal to the site of pathology will be normal. Similarly, the CMAP amplitude will be normal if the stimulation site is below the demyelinated segment. For that reason, in any suspected demyelinating mononeuropathy, an attempt should be made if technically possible to stimulate the nerve in question above and if at all possible across (inching) the affected site. This has the benefit of not only identifying the existence of the abnormality, but also precisely localize it, while at the same time providing valuable prognostic information. Needle examination findings in a demyelinating mononeuropathy consist only of reduced recruitment of normal appearing MUAPs and then only if the pathophysiology is that of conduction block. Focal slowing and temporal dispersion are not associated with abnormalities on needle examination. As many nerve injuries include demyelinating and axonal components, it is not uncommon to identify EDX features associated with both types of injuries.

IMAGING

Historically, x-rays of the lumbosacral spine were performed routinely in patients with back or radicular pain. In consideration of radiation exposure and their very limited yield in this clinical context, we agree with those who would utilize routine back x-rays for those with significant trauma, those with symptoms or at high risk for systemic disease, or those with histories suggesting recent compression fracture.⁵ When indicated and when feasible, magnetic resonance imaging (MRI) imaging is the imaging procedure of choice for suspected mono- or polyradiculopathy. Although we have a low threshold for ordering MRIs in individuals with radiculopathy and neurologic deficits, we do not consider it mandatory. We are comfortable following someone clinically when all information points to a routine compressive radiculopathy due to disc herniation, as long as improvement with subsequent evaluations can be demonstrated. Information gleaned from imaging studies requires careful clinical correlation as incidental findings are exceedingly frequent. Depending on age, herniated discs are identifiable in 20–40% of asymptomatic individuals.⁵ Bulging discs are even more prevalent, identifiable in up to 80% of asymptomatic volunteers.⁵ The decision to utilize gadolinium is individualized. It provides limited benefit in typical discogenic or spondylotic disease and poses some risk, particularly in those with reduced glomerular filtration. Gadolinium is most likely to be helpful in those with prior back surgery or when there is suspicion of systemic disease as a cause of radiculopathy. When MRI is precluded for any reason, post-myelographic CT scans provide an excellent imaging surrogate for nerve root disease. Although ultrasound appears to have an increasing role in neuromuscular disease, it is felt to be of limited or no utility in the evaluation of radiculopathy.¹⁶

MRI has become the modality of choice for evaluation of lumbosacral plexopathy or radiculoplexopathy as well, particularly utilizing 3 T or higher neurography techniques.¹⁷ It is of value not only with structural pathology of nerve like neurofibromatosis but is of benefit in presumed inflammatory, ischemic injury in disorders such as non-diabetic lumbosacral radiculoplexus neuropathy (non-DLRPN).^17,18 Its resolution is in general superior to CT, and it provides the added benefit of readily providing axial, sagittal, and coronal viewing planes. Gadolinium is often of value, as neoplastic and inflammatory conditions are both relatively common causes of plexopathy whose visualization and characterization will be enhanced with the addition of gadolinium.¹⁹

Imaging of mononeuropathies in the lower extremities, particularly in proximal locations, is rendered difficult by the small diameter and circuitous course of the nerves in conjunction with the complex anatomy of the region. MRI and ultrasound imaging of at least seven nerves (femoral, lateral femoral cutaneous, obturator, sciatic, superior and inferior gluteal, and pudendal) are feasible and warranted in the case of unexplained or progressive neuropathies identified by clinical and/or EDX means.¹⁰ The imaging may allow for identification of focal T2 signal abnormalities at sites of compression, nerve enlargement and enhancement from neural tumors, enhancement from focal inflammatory lesions, or external compression from any contiguous mass.¹⁷ Imaging may also have a therapeutic application, allowing for more precise application of steroid injections, for example, in obese individuals with meralgia paresthetica in whom normal anatomic landmarks may be difficult to identify.¹⁰ Recently, diffusion tensor imaging has been utilized axonal changes in patients with peripheral neuropathy. Conceivably, this or similar technologies may provide the ability to image and monitor axonal regrowth subsequent to injury and potential therapeutic intervention.²⁰

MONORADICULOPATHIES

Lumbosacral radiculopathies are more prevalent than their cervical or thoracic counterparts. For the most part, they occur as a consequence of mechanical compression from some aspect of spondyloarthropathy that is narrowing of the central canal or lateral recesses, and/or neural foramina by disc material, osteophyte, hypertrophied ligament or some combination thereof. Less commonly, they may result from compression from benign or malignant neoplasm, hematoma or abscess. They may result as well from neoplastic, infectious and inflammatory disorders with a predilection to attack or invade nerve roots or meninges (Table 24-3 and Fig. 24-10). A heightened index of suspicion is required for these less common causes. The primary symptom of monoradiculopathy is pain, commonly described as radicular or sciatic due to its linear trajectory following the course of the sciatic nerve in most cases. It has been estimated that disc herniations associated with objective neurologic deficits occur in the complete absence of pain in only 1/1,000 patients.⁵

Radicular pain in the lower extremity often occurs in the absence of significant back pain and often begins in the sacroiliac or gluteal regions. Although commonly continuous, it may be interrupted, for example, affecting the buttock and anterior leg but skipping the thigh. The pain is often positional depending on the exact site and vector of compression, often related to specific back postures which may increase or decrease the cross-sectional area of the central canal or neural foramina. Limb pain that is aggravated by side-bending toward the affected side or by straight leg raising of the ipsilateral or contralateral leg is likely to be due to nerve root compression. Radicular pain induced by straight leg raising of less than 60 degrees is a sensitive but nonspecific sign estimated to occur in 90% of patients with radiculopathy secondary to disc herniation.²¹ Conversely, reproduction of ipsilateral radicular pain by raising the opposite limb, that is, reverse straight leg raising, is a highly specific but fairly insensitive provocative test.⁵ In upper lumbar disc disease, pain may be reproduced by reverse straight raising, that is, by passively extending rather than flexing the thigh at the hip joint.²²

Pain may also be increased by increased pressure within the intraspinal canal. The latter is frequently provoked by maneuvers that increase intrathoracic pressure resulting in increased volume of the epidural venous plexus. Pain radiating down the leg provoked by straining or coughing is therefore a helpful although inconsistent clinical clue.

Regarding the examination of a patient with neurologic complaints of the lower extremity, there are a number of notable differences in comparison to the upper limb. The lower extremity has fewer testable muscles and actions than the upper extremity. For example, there is no pronation or supination at the knee, and the testable options of the toes is limited in comparison to the fingers. In the lower extremities, there may be greater difficulty in distinguishing a nerve from a nerve root lesion as there is greater overlap in both the motor and sensory functions of specific nerves and nerve roots. For example, there are more similarities than differences in the motor, sensory, and reflex findings in an L3–4 radiculopathy and femoral neuropathy. One advantage that the lower extremity holds over the upper extremity however, both clinically and electrodiagnostically, is that that muscles belonging to the same myotome can be found in both proximal and distal locations. For example, the L5 segment contributes significantly to both toe extension and hip abduction, whereas the C5 segment has no meaningful contribution to distal upper extremity functions such as wrist or finger movement.

It is also important for a clinician to recognize that clinically evident motor deficits in monoradiculopathy may be subtle if evident at all due to the typical multisegmental innervation of virtually all muscles. As a corollary of this, weakness from monoradiculopathy when present should not produce complete paralysis of any muscle. It is also important to recognize that the pattern of weakness in monoradiculopathy although segmental, typically does not affect all muscles innervated by a single myotome to the same extent (Table 24-4). This is particularly true for muscles that are more proximally located in a given segment. For example, the weakness in an L5 radiculopathy may be confined to great toe extension and is infrequently detected in hip abduction. Relevant deep tendon reflexes in monoradiculopathies are typically reduced or are absent in the affected segment.

As previously emphasized, sensory symptoms in monoradiculopathy, like many neurologic diseases, are a more sensitive indicator of sensory involvement than sensory signs. Again, with suspected L5 radiculopathy, a complaint of a numb big toe should be considered as a valid complaint even in the absence of convincing sensory deficits on examination. In addition, it is not uncommon for paresthesias and sensory loss to persist long after radicular pain and demonstrable weakness have resolved. Another potential source of error in the interpretation of radicular sensory involvement is the failure to recognize that the topographical area of sensory involvement described by the patient or demonstrable on examination is typically far smaller than predicted on the basis of the commonly published dermatomal maps (Fig. 24-11).

Although the vast majority of compressive monoradiculopathies have pain as their cardinal symptom at some point in their natural history as described above, it is important to recognize that there appear to be exceptions. Patients may have either dermatomal sensory symptoms or myotomal motor signs in the absence or relative absence of pain. On occasion, patients will have radicular pain with paresthesia that will abruptly resolve, only to be replaced by weakness in a segmental pattern. It has been hypothesized that this may occur as a result of disc sequestration and migration.^23,24 The authors’ personal observations of this syndrome occurring in the immediate aftermath of spinal manipulation would support this contention.

L1–2

Monoradiculopathies affecting these roots are uncommon and typically present with pain referred into the inguinal region and perhaps the proximal, anterior thigh. Pain in this region is however, uncommonly neurogenic in nature. Other than L1–2 radiculopathies, neurogenic pain with this topographic distribution may result from mononeuropathies of the ilioinguinal or genitofemoral nerves. Although discogenic L1/L2 radiculopathy may occur, suspected L1/L2 radiculopathy should generate an increased level of suspicion for an unusual etiology of root disease (Fig. 24-12). Paresthesias occur in the trochanteric and/or upper groin regions in L1 lesions and the anterior thigh in L2. Weakness is uncommon but may be detectable in hip flexion with L2 root disease. The ipsilateral cremasteric reflex may be lost.

L3–4

L3–4 monoradiculopathies have the potential for substantial morbidity if quadriceps weakness occurs. Pain and sensory symptoms of the thigh and medial knee imply L3 involvement, whereas involvement of the medial leg implicates the L4 root. Either lesion may lead to a diminished or absent knee jerk and weakness of hip flexion and adduction in addition to the critical function of knee extension. As the quadriceps is a particularly strong muscle, mild weakness may only be detectable in a younger, nonobese patient by asking the patient to get up from a chair on one leg at a time without using the arms. Alternatively, mild weakness may be detected by asking the patient to do a partial squat while weight bearing on one leg alone. Both tests should be applied cautiously. In either case, the examiner must position themselves in such a manner and be confident that they can support the patient and prevent a fall should one occur with either maneuver. Many texts suggest that the TA receives partial innervation through the L4 segment. In the authors’ experience and in the published experience of others, weakness of foot dorsiflexion and denervation of the TA rarely occur in documented L4 monoradiculopathies.¹⁵ The differential diagnosis of L3–4 radiculopathy includes femoral mononeuropathies, lumbar plexopathies, and radiculoplexus neuropathies. Clinical and electrodiagnostic sparing of hip abductors distinguishes a femoral mononeuropathy from any of the other disorders. The more difficult distinction is from lumbar radiculoplexus neuropathies or plexopathies which may share a similar pattern of pain, sensory involvement, weakness, and denervation. A combination of imaging of the back and retroperitoneum, EDX, and the clinical contextual features may be required to resolve this differential diagnostic dilemma.

L5

L5 is the most common lower extremity monoradiculopathy. The pain typically extends from buttock to posterolateral thigh to anterolateral leg. Sensory symptoms are felt in the lateral leg, instep, dorsum of the foot, and particularly in the big toe. In most people, weakness will be most commonly and readily detected in great toe extension. Weakness of foot dorsiflexion and typically inversion > eversion may occur as well. Weakness of knee flexion and hip abduction are less frequent and/or are more difficult to detect. The differential diagnosis of L5 radiculopathies is essentially the differential diagnosis of foot drop. As many causes of polyneuropathy, motor neuron disease, and even myopathy have a predilection to affect foot dorsiflexion, this differential diagnosis is expansive. As the most commonly occurring of these is a common fibular neuropathy, it is extremely important to assess the strength of ankle inversion which should not be affected in a common fibular (peroneal) neuropathy.

As suggested, a diagnosis of L5 monoradiculopathy should be made cautiously in the absence of pain and/or sensory symptoms, when foot dorsiflexion weakness exceeds toe extension weakness or when the ipsilateral ankle jerk is depressed. The former raises the possibility of motor neuron disease, rarely myasthenia or distal myopathy particularly if bilateral and symmetric. Foot drop greater than or without toe drop suggests distal myopathy or upper motor neuron disease. An absent ankle jerk should not occur in an L5 monoradiculopathy and suggests more proximal pathology of the sciatic nerve in which the fibular nerve is more severely affected, or concomitant involvement of the S1 nerve root.

S1

This is the second most common lower extremity monoradiculopathy. The radicular or “sciatic” pain of S1 root disease typically extends from buttock down the posterior surface of thigh and leg into the heel and at times into the lateral toes. Sensory symptoms are most pronounced in the posterior lateral leg and particularly in the lateral and plantar surfaces of the foot and little toe. Muscle weakness, if present, most commonly occurs in foot plantar flexion. Detection of S1 weakness may be hampered by the considerable baseline strength of foot plantar flexion, knee flexion, and hip extension. Functional testing, for example, the ability to elevate the heel fully off the floor while standing on one leg alone while keeping the knee fully extended is helpful in this regard. As this is a test of strength, not balance, the patients should be supported by either their examiner or a nearby wall. Detectable weakness in S1-innervated hip extension is uncommon and is most rigorously tested by having the patient extend the thigh at the hip while in the prone position. A suppressed or absent ankle jerk is expected.

The differential diagnosis of an S1 radiculopathy is not as extensive as its L5 counterpart. Tibial, sural, and plantar neuropathies are uncommon. Sciatic neuropathies are more common. Most cases are dominated however by deficits arising from its fibular division. Sacral plexopathies are usually readily distinguished from S1 root disease by the more widespread pattern of weakness and sensory loss.

Table 24-4 lists the more common causes of monoradiculopathies. Herniation of intervertebral discs causes the vast majority of monoradiculopathies in those younger than 50 years. Spondylosis is a far more common cause of root disease in older adults. Typically, the clinical deficits are more evident in a monoradiculopathy caused by disc herniation than in spondylosis, presumably due to its relative acuity. Spondylotic disease is typically more insidious in its development and commonly affects multiple levels bilaterally in an older population, a pattern that may be evident only through electrodiagnostic evaluation. Contrary to common belief, disc herniations are rarely the result of significant traumas such as motor vehicle accidents. Nonspondylotic causes of monoradiculopathy deserve a higher index of suspicion in individuals at risk (e.g., individuals who are immunosuppressed or with prior history of malignancy); those with fever, weight loss, or other symptoms of systemic disease; or those whose neurologic deficits progress.

The differential diagnosis of monoradiculopathy is limited. Neoplasms, either nerve sheath tumors or malignancies affecting vertebrae or meninges are one cause of nondiscogenic/spondylytic monoradiculopathy (Fig. 24-10). Any slowly progressive monoradiculopathy should prompt a careful discussion of family history and search for neurocutaneous stigmata, including subcutaneous and Lisch nodules, café-au-lait spots, and axillary freckles. Diabetes is not commonly considered as a cause of monoradiculopathy, but self-limited monoradiculopathies will occasionally occur in diabetics in the absence of other apparent causes. Herpes zoster (shingles), is a fairly common cause of radicular pain and sensory loss but on occasion can produce “zoster motor paresis” in approximately 5% of affected individuals, presumably due to retrograde viral movement from the dorsal root ganglia into the anterior horn.^25–30 Other disorders with an affinity for nerve roots may occasionally present as a monoradiculopathy as an initial manifestation of an evolving polyradiculopathy (Table 24-5).