Anatomy and Physiology of Peripheral Nerves

The anatomic site of involvement may range from the nerve cell body (neuronopathy, e.g., acute ataxic sensory neuronopathy), to the nerve roots (acute inflammatory demyelinating polyradiculoneuropathy [AIDP]), to the nerve proper, to the axon terminals (twig neuropathy, e.g., hyperparathyroidism). Nerve fibers react to injury through two primary mechanisms, axonal degeneration and demyelination. In axonal degeneration the cytoplasm disintegrates. Eventually, the axolemma becomes discontinuous, and the myelin sheath, but not the Schwann cell, secondarily disintegrates. Fragments of axolemma and myelin form linear arrays of myelin ovoids. Macrophages phagocytose and eventually remove the debris. Conditions affecting the axon typically begin in the most terminal axon twigs, and the pathologic changes move proximally with progression of the disease, a phenomenon known as dying back. Clinically, signs and symptoms begin in the longest axons, those innervating the feet and toes, and progressively involve more proximal areas, a feature known as length dependency. Wallerian degeneration specifically refers to axonal degeneration distal to a traumatic nerve injury.

In demyelinating neuropathies, the primary insult is to the myelin sheath or the Schwann cell, and varying degrees of interference with myelin maintenance are produced. In the earliest stages, the demyelination is paranodal, but with progression, the rest of the internodal segment becomes involved. Segmental demyelination refers to the involvement of several internodes over a segment of nerve. In some conditions the Schwann cells are involved diffusely and the demyelination is uniform along the length of the nerve (e.g., Charcot-Marie-Tooth [CMT] disease). In most acquired myelinopathies, the involvement is multifocal (e.g., Guillain-Barré syndrome [GBS]). In compression neuropathies, the demyelination is focal and involves only a discrete segment (e.g., carpal tunnel syndrome).

With severe demyelination, there may be secondary axon loss. In compression neuropathies, the myelin is more susceptible to injury and axon loss reflects chronicity and severity. In dysimmune inflammatory neuropathies, the axon may be an accidental victim of the attack, the bystander effect. The secondary axon loss in myelinopathies and the secondary demyelination in axonopathies add another measure of complexity to the evaluation of neuropathies. Clinically and electrophysiologically, these secondary changes may create a mixed picture of axonopathy and myelinopathy. For example, uremia characteristically produces an axonopathy with secondary demyelination, and severe GBS, a primarily demyelinating and inflammatory neuropathy, is often associated with significant secondary axon loss as a result of the bystander effect.

With the use of clinical, electrodiagnostic, pathologic, and laboratory data, neuropathies can be classified according to their anatomic site of involvement, time course, distribution, pattern of deficit, type of functional fiber involved, and whether the process is primarily axonal, primarily demyelinating, or mixed. It is helpful to classify a neuropathy in as many ways as possible in the hope of uncovering a characteristic signature. There are two primary patterns of clinical involvement: generalized, diffuse, or symmetrical versus focal or multifocal. The time course may vary from acute, evolving over a period of days to weeks, to chronic, evolving over a period of months to years, to very chronic or lifelong, to relapsing and remitting. Neuropathies may be predominantly motor (e.g., GBS) or predominantly sensory (e.g., amyloidosis). Most neuropathies produce symmetrical distal involvement, but some may cause primarily proximal weakness (e.g., porphyria), proximal sensory loss (e.g., Tangier disease), or acral sensory loss over the coolest parts of the body (e.g., leprosy). Sensory neuropathies may preferentially affect large fibers (e.g., vitamin B₁₂ deficiency) or small fibers (e.g., hereditary sensory neuropathy). Particular neuropathies may or may not have associated autonomic dysfunction. As defined by clinical features, by electrodiagnostic profile as determined by nerve conduction studies and needle electromyography (EMG), and occasionally by pathologic examination, a neuropathy may be primarily axonal (e.g., CMT type 2 [CMT2]), primarily demyelinating (e.g., chronic inflammatory demyelinating polyradiculoneuropathy [CIDP]), mixed but mainly axonal (e.g., chronic renal disease), or mixed but with significant demyelination (e.g., diabetes).

To establish an etiologic diagnosis, the neuropathy must be characterized and classified as precisely as possible. To separate the cause of a particular neuropathy from the protean possibilities, an attempt should be made to construct a profile of clinical and electrophysiologic characteristics to narrow the differential diagnosis. From the long list of causes of neuropathy, only a short list of conditions emerges that cause a chronic, predominantly sensory, large-fiber axonopathy or an acute, predominantly motor, multifocal demyelinating neuropathy. Knowledge of the relevant anatomic, physiologic, and pathologic concepts helps greatly in this exercise.

Clinical Signs and Symptoms of Peripheral Neuropathy

The cardinal manifestations of peripheral neuropathy are weakness, alterations in sensation, and changes in reflexes. The patterns of abnormality are important in the differential diagnosis, and detailed clinical examination in which the details and specifics are explored is often rewarded. The first step in the differential diagnosis is determining the pattern of motor and sensory dysfunction. Is the neuropathy predominantly motor or sensory? If weakness has occurred, is it proximal, distal, or diffuse? Is atrophy present? Are the changes in reflexes dependent on length or are they global? Do the sensory changes demonstrate a predilection for large or small fibers? Is there autonomic dysfunction? Are the findings symmetrical?

Generalized, length-dependent peripheral neuropathies are typically symmetrical in onset, with numbness beginning initially in the toes and feet and affecting the lower part of the legs as the neuropathy progresses. The fingers and hands begin to be affected about the time that lower extremity involvement reaches the knees. This reflects the length-dependent nature of the process because nerve length from the lumbosacral dorsal root ganglia to the knees is comparable to the length from the cervical dorsal root ganglia to the fingers. The most terminal distribution of the intercostal nerves is a strip along the chest and upper part of the abdomen, and in a length-dependent process, sensory loss can sometimes be found in this region, referred to as a cuirass distribution. With motor involvement, weakness typically begins in the toe flexors and extensors, followed by the development of footdrop. The foot plantar flexors are so powerful that it is seldom possible to find weakness unless the neuropathy is very severe. In very chronic neuropathies, it is common to see high arched feet and hammer toes because the greater strength of the extrinsic than the intrinsic foot muscles in the lower part of the leg pulls the toes into this position. Obvious atrophy of the intrinsic muscles with ridging between the long toe extensor tendons is often present. In a length-dependent process the first tendon reflexes to disappear are the ankle jerks; as the disease progresses, knee jerks are lost at about the same time as the upper extremity reflexes begin to be depressed because of the similar length of their axons.

In the differential diagnosis, age at onset, pace of evolution, presence of associated medical or neurologic conditions, and medication history are very important. The family history is crucial but often tricky because many individuals with hereditary neuropathies do not recognize their own condition. Other family members with odd-looking feet (pes cavus, characteristic of CMT disease) or feet that are difficult to fit with shoes may be the only clue to a condition that is rampant but unrecognized in a kindred.

Quantitation of the degree of deficit is useful for longitudinal follow-up. Sophisticated instrumentation is available for this purpose but is expensive and not in wide use outside academic centers. Simple bedside testing can provide a great deal of useful information. Strength is most often described according to the Medical Research Council scale. The power in small hand muscles can be nicely described with techniques that compare the patient’s strength with the examiner’s.² Two-point discrimination is quantitative and easily performed. Inexpensive handheld esthesiometers can quantitate touch sensibility. Vibratory sensation is a sensitive parameter of peripheral nerve function and can be simply quantitated by noting where the patient can perceive the sensation and for how long (e.g., “vibration absent at the great toes and metatarsal heads, can perceive a maximally vibrating 128-Hz fork for 5 seconds over the medial malleoli”). If the patient returns and is found to have lost vibration sensibility over the malleoli, the condition is progressing. If on follow-up vibration sensation is present for 12 seconds over the malleoli and can now be perceived for 3 seconds over the metatarsal heads, the patient is improving. Functional testing is invaluable. The time required to walk a set distance or get up from the floor, arm and leg outstretch time, the ability to support the entire body weight on one tiptoe, and similar functions are objective and quantifiable.

The onset of neuropathy may be relatively acute or subacute and evolve over a period of several days to several weeks or be more chronic and evolve over a period of several months to many years. There is some unavoidable overlap in the middle, between what might be called subacute and what might be called chronic. Some neuropathies have a predilection for certain types and sizes of fiber. Large-fiber neuropathies affect strength, reflexes, and proprioception, with relative sparing of pain and temperature sensation, whereas small-fiber neuropathies primarily affect pain, temperature, and autonomic function, often with strength and reflexes being spared. Differential involvement of large versus small sensory fibers can sometimes be discerned clinically. Standard nerve conduction studies evaluate the conduction characteristics of only large, myelinated fibers.

Electrophysiology of Peripheral Neuropathies

Demyelinating neuropathies and axonopathies generate quite different electrophysiologic pictures. In brief, primary myelinopathies produce marked slowing of conduction velocity (CV) with preservation of distal compound muscle action potential (CMAP) amplitudes, whereas axonopathies produce marked loss of distal amplitudes with relative preservation of CV. When the disease process affects only myelin, conduction is impaired because of loss of saltatory conduction. If the process is uniform and diffuse with all myelin being affected, as in a hereditary myelinopathy, the conduction abnormalities involve all nerves and all segments of nerves to an approximately equal degree. If the process is focal, only the involved nerve is affected. If the process is multifocal, the abnormalities may be widespread but manifested to different degrees in different nerves and in different segments of the same nerve. In either case, with a myelinopathy the axon is intact and there is no loss of trophic influences on the target motor and sensory organs. Muscle atrophy does not occur, and denervation potentials, fibrillations, and positive sharp waves do not appear in the affected muscles.

In contrast, a disease process that primarily affects axons also causes loss of the axon-mediated trophic influences on target organs. The muscle becomes atrophic, with a resultant decline in CMAP amplitude. Axon loss produces a diminution in sensory or compound nerve action potential amplitude. Involvement of all the axons in a nerve may render the nerve inexcitable, and no motor or sensory potentials can be elicited. However, any surviving axons will conduct at their normal velocity. If a disease process produces dropout of 50% of the axons in a nerve, the electrodiagnostic picture would be a 50% (Å) loss of CMAP amplitude but relatively normal CV. Clinical and electrophysiologic characteristics helpful in distinguishing axonopathy from myelinopathy are summarized in Table 233-2.

TABLE 233-2 Clinical and Electrodiagnostic Features That Help Distinguish Axonopathy from Myelinopathy

CMAP, compound muscle action potential; CSF, cerebrospinal fluid; CV, conduction velocity; DML, distal motor latency; NAP, nerve action potential; SNAP, sensory nerve action potential.

Unfortunately, as is often the case, findings do not always follow the classic or typical pattern. Some neuropathies are truly mixed, with electrophysiologic features of both demyelination and axon loss. In addition, axonopathies can produce some degree of slowing of CV. Random involvement of axons will inevitably affect some of the largest and fastest conducting fibers, whose dropout will then lower the maximum CV. If the largest and fastest conducting axons are preferentially affected, significant slowing could occur as a result of an axonopathy. The severity of an axonopathy is reflected by the degree of muscle fiber atrophy and its attendant loss of CMAP amplitude. A severe axonopathy can be expected to involve at least some of the fastest conducting fibers and will therefore produce more CV slowing than will a mild axonopathy. For this reason, if the CMAP amplitude is decreased, more severe conduction slowing must be present before one can be confident of the existence of demyelination. Because of these complicating factors, criteria for demyelination have been developed. Despite the lack of universal consensus, the use of some criteria set is coming into increasingly wide acceptance, but debate continues regarding precise details.

Although CV and CMAP amplitude are the primary parameters, other electrodiagnostic variables may be useful as well. Disproportionate prolongation of distal motor latency or late response latency, beyond that explicable on the basis of axon loss, may also indicate demyelination. Axonopathies are length dependent and exhibit a “dying-back” process that affects the most distal nerve terminals first and involves more proximal nerve segments with progression. Disproportionate conduction abnormalities in the most distal nerves can thus indicate axonopathy. A gradient of abnormality on needle examination, with greatest involvement of the most distal muscles and progressively less involvement of more proximal muscles, suggests a length-dependent process.

Another major indicator of demyelination is conduction block or temporal dispersion. This is a change in amplitude or configuration of the CMAP waveform on stimulation proximal to a given point. In temporal dispersion, focal demyelination of axons produces CV slowing that involves some fibers to a greater degree than other fibers and thereby causes loss of synchrony. The CMAP conducted through the demyelinated region is dispersed and spread out, with an increased duration and loss of amplitude. However, all fibers successfully conduct through the affected region, so the total area under the CMAP curve remains the same with both proximal and distal stimulation. In conduction block, some fibers are so severely demyelinated that they do not conduct at all. There is not only a diminution in CMAP amplitude but also a diminution in total area under the curve, thus indicating that fewer muscle fibers have been successfully depolarized with proximal stimulation than with distal stimulation.

Clinical weakness seems to correlate with conduction block, not with CV slowing or temporal dispersion. A great deal has been made of distinguishing between conduction block and temporal dispersion, but the distinction is of questionable utility because both these phenomena indicate a focal demyelinating lesion in the subjacent nerve.³ As a result, the fairly simple criterion of a significant loss of negative spike or peak-to-peak amplitude with proximal stimulation as compared with distal stimulation may suffice to indicate demyelination, and whether the loss of amplitude is due to conduction block, temporal dispersion, or a combination of the two is relatively immaterial.

Demyelination may occur in two major patterns. Disorders that affect myelin diffusely because of a genetic defect, biochemical abnormality, or the effect of certain drugs or toxins produce global, uniform demyelination. There is little variation from nerve to nerve or from segment to segment of any given nerve.⁴ In the majority of patients with familial demyelinating neuropathy, distal motor latencies are prolonged in proportion to CV, median and ulnar forearm CV does not vary by greater than 5 m/sec, and there is no evidence of conduction block or temporal dispersion. Such uniform slowing of conduction suggests a generalized dysfunction of myelin or Schwann cells.

In acquired demyelinating neuropathies, distal motor latency and CV vary more randomly, differences between median and ulnar CV of greater than 5 m/sec or even 10 m/sec are common, and conduction block or temporal dispersion may occur. Such a pattern of conduction abnormality suggests a multifocal attack on peripheral nerves that may become widespread but does not truly affect myelin diffusely.

Other laboratory tests are often useful. Macrocytosis on the hematology profile may be a clue to vitamin B₁₂ deficiency or to alcohol abuse. If alcoholism is strongly suspected, carbohydrate-deficient transferrin may be useful. It reflects the level of alcohol intake over the preceding weeks or months in the same way that glycosylated hemoglobin reflects chronic blood sugar levels. An elevated carbohydrate-deficient transferrin level may be a clue to occult alcohol abuse as the cause of a generalized polyneuropathy.⁵ Abnormal liver function test results, especially γ-glutamyltransferase, may reflect alcohol abuse or a neuropathy related to underlying hepatitis, especially hepatitis C. Hepatitis C may cause a neuropathy associated with cryoglobulinemia. Cryoglobulins are difficult to detect; they are large complex molecules that may cross-react with rheumatoid factor, so a positive rheumatoid factor test may be a clue to hepatitis C. Many other systemic diseases may be manifested as a peripheral neuropathy, including connective tissue disorders, systemic vasculitis, vitamin B₁₂ deficiency, Lyme disease, paraproteinemias, porphyria, hypothyroidism, human immunodeficiency virus (HIV) infection, amyloidosis, sarcoidosis, and occult malignancy.

In GBS and CIDP, cerebrospinal fluid (CSF) protein is often elevated, and lumbar puncture should be performed whenever these conditions are in the differential diagnosis. In CIDP, magnetic resonance imaging of the lumbosacral spine may reveal enlarged and inflamed lumbosacral roots. Formal autonomic testing may help establish whether a neuropathy has a component of dysautonomia. In recent years, skin biopsy has emerged as a safe, minimally invasive tool for assessing small epidermal nerve fibers that are inaccessible for routine neurophysiologic tests. Biopsy of hairy skin can be used to evaluate unmyelinated and thinly myelinated fibers, and biopsy of glabrous skin can be used to examine large myelinated fibers. Standard morphometric techniques have been developed and proven to be reliable and reproducible. Pathologic changes in cutaneous nerves have been found to occur very early in the course of peripheral neuropathies. Comparison of the density of the epidermal nerve fiber layer between a proximal and a distal biopsy site can sometimes document the length dependency of a neuropathy, as well as occasionally make the diagnosis in a condition such as amyloidosis.

Peripheral neuropathy is obviously common in patients with diabetes mellitus. Recently, controversy has arisen regarding the diagnosis of neuropathy in patients with an abnormal 2-hour glucose tolerance test but without frank diabetes and the proper screening test for patients with idiopathic peripheral neuropathy, especially sensory neuropathy. Undiagnosed impaired fasting glucose metabolism appears to be associated with neuropathy at a higher frequency than in the general population when the 2-hour oral glucose tolerance test is used as opposed to a fasting blood sugar and glycosylated hemoglobin level.⁶ In a study of 100 consecutive patients with apparently idiopathic peripheral neuropathy, the prevalence of undiagnosed abnormal fasting glucose metabolism was found to be nearly twofold higher (62%) in patients with neuropathy than in controls. The 2-hour oral glucose tolerance test provided a higher diagnostic rate with the 2003 revised American Diabetes Association criteria. There is increasing evidence that abnormal glucose metabolism, or prediabetes, may be a risk factor for neuropathy and that a 2-hour oral glucose tolerance test may be more sensitive than fasting blood sugar and glycosylated hemoglobin in detecting this condition.

In Tables 233-3 to 233-6, neuropathies are classified into demyelinating versus axonal and whether the onset is acute/subacute versus subacute/chronic. Other common classifications include (1) mixed axonopathy/myelinopathy: diabetes, end-stage renal disease (ESRD), and some cases of acute or chronic inflammatory demyelinating neuropathy; (2) primarily motor polyneuropathies: most cases of GBS, CIDP, porphyria, dapsone and lead intoxication, and CMT disease; (3) large-fiber sensory neuropathies: uremia, diabetes (pseudotabes), paraneoplastic neuropathy, Sjögren’s syndrome, vitamin B₁₂ deficiency, Friedreich’s ataxia, certain toxins (e.g., pyridoxine, cisplatin, and metronidazole), acute idiopathic sensory neuronopathy, and some cases of CIDP; (4) small-fiber sensory neuropathies: diabetes mellitus (pseudosyringomyelia), amyloidosis, hereditary sensory autonomic neuropathies, and leprosy; and (5) neuropathies with major autonomic dysfunction: diabetes, alcoholism, amyloidosis, hereditary sensory autonomic neuropathy type III (Riley-Day syndrome), GBS, porphyria, vincristine toxicity, and idiopathic pandysautonomia.

TABLE 233-3 Acute/Subacute Myelinopathies

* Both patterns have been reported.

† Recall that Guillain-Barré syndrome is a syndrome, not a disease, and can occur as part of several conditions.

TABLE 233-4 Subacute/Chronic Demyelinating Neuropathies