In our experience and others, a cause for neuropathy will not be found in as many as 50% of cases despite an extensive work-up.^1–11 The chronic idiopathic polyneuropathies are likely a heterogeneous group of neuropathies. Most individuals have only sensory symptoms, but some may have mild weakness (e.g., toe extension) or slight abnormalities on motor conduction studies. The neuropathy may affect large- and/or small-diameter nerve fibers. As the etiology is unknown, only symptomatic management of the neuropathic pain is available.

CLINICAL FEATURES

Most individuals present with numbness, tingling, or pain (e.g., sharp stabbing paresthesias, burning, or deep aching sensation) in the feet between the ages of 45 and 70 years.^1–11 This is a common problem occurring in approximately 3% of adults as they age. In a large series of 93 patients with idiopathic sensory polyneuropathy, 63% presented with numbness and paresthesia along with pain, 24% with numbness or paresthesia without pain, and 10% with pain alone.⁹ Eventually, 65–80% of affected individuals develop neuropathic pain.^6,9–11 Sensory symptoms are first noted in the toes and slowly progress up the legs and later into the arms. The average time to involvement of the hands is approximately 5 years.^6,9

Neurological examination reveals the typical length-dependent pattern of sensory loss.^6,7,9,11 Vibratory perception is reduced in 80–100%, proprioception is impaired in 20–30%, pinprick sensation is diminished in 75–85%, and light touch is decreased in 54–92% of those with the neuropathy. Strength is usually normal, although mild distal weakness and atrophy involving toe muscles may be appreciated in 40–75% of cases, and rarely of ankle dorsiflexors and plantar flexors.^6,9,11 However, upper limb strength, including the hand intrinsics, should be normal. Muscle stretch reflexes are usually absent at the ankle and diminished at the knees and arms. Generalized areflexia though is less common and would point to a hereditary or acquired demyelinating neuropathy.

Within the category of idiopathic sensory or sensorimotor polyneuropathies are people who have only a small fiber sensory neuropathy.^2,3,7,9 By definition, these individuals should have normal nerve conduction studies (NCS), and nerve biopsies, if performed, demonstrate a relatively normal density of large myelinated nerve fibers. Most people with small fiber neuropathy (approximately 80%) complain of burning pain in the feet, while 40–60% describe sharp, lancinating pain; paresthesias; or just numbness. Symptoms may involve the distal upper extremities. Rarely, the neuropathy is restricted to the arms and face or involves the autonomic nervous system.^2,3 Examination reveals reduced pinprick or temperature sensation in almost all patients, while vibratory perception is impaired in half. Muscle strength is preserved. Likewise, muscle stretch reflexes are also usually normal, but a few patients have reduced reflexes at the ankles.

LABORATORY FEATURES

The diagnosis of chronic idiopathic polyneuropathy is one of exclusion. Laboratory testing should include fasting blood glucose (FBS), hemoglobin A1 C (HgbA1 C), antinuclear antibody, anti-Ro and anti-La antibodies (SSA and SSB), erythrocyte sedimentation rate, B12, serum and urine immunoelectrophoresis/immunofixation, and thyroid, liver, and renal function tests.^12,13 If the FBS and HgbA1 C are normal, we typically order an oral glucose tolerance test (GTT). The most common abnormality, when one is found, in patients with sensory neuropathy is diabetes or impaired glucose tolerance (IGT). IGT (defined as glucose of >140 and <200 mg/dL on 2-hour GGT) is seen in 17–61% and frank diabetes mellitus (DM) (defined as 2-hour glucose of >200 mg/dL on GGT or FBS of >126 mg/dL) in 20–31% of patients with sensory neuropathy (Table 22-1).^14–18 In patients with painful sensory symptoms (not just numbness), the likelihood of IGT or DM is even higher. However, some authorities have not found increased risk of IGT in their patients with idiopathic neuropathy compared to age-matched controls.¹⁹ Thus, although the risk of both previously undetected DM and IGT may be increased in patients with sensory neuropathy, this is still controversial and a causal relationship has not been firmly established.^20,21

About 5% of patients with chronic idiopathic sensory or sensorimotor polyneuropathy have a monoclonal protein detected in the serum or urine, but this is not much higher than the age-matched normal controls. Furthermore, the relationship of these monoclonal proteins to the pathogenesis of most neuropathies is unclear. There is a strong pathogenic relationship established in people with demyelinating sensorimotor polyneuropathies with IgM monoclonal proteins, half of whom have myelin-associated glycoprotein (MAG) antibodies (discussed in Chapters 14 and 19). However, most individuals with chronic idiopathic sensory or sensorimotor polyneuropathy have axonal neuropathies both histologically and electrophysiologically. Amyloidosis is the other condition in which a pathogenic relationship between the neuropathy and the monoclonal protein is clear. Thus, amyloid neuropathy needs to be excluded in patients with a monoclonal gammopathy before concluding that the neuropathy is idiopathic in nature (see Chapter 16). This may require a fat pad, rectal, bone marrow, or nerve biopsy.

Although some studies have suggested that antisulfatide antibodies are common with painful small fiber neuropathy,^22,23 subsequent reports suggest that these antibodies have a very low sensitivity and poor specificity.^6,10 We never order them as we have found them to be of little use clinically, and a pathogenic relationship has never been demonstrated. That is, the presence of these antibodies does not imply that the patients have an immune-mediated neuropathy and that they may respond to treatment with immunotherapy. We also feel that there is no role for screening various antiganglioside and other antinerve antibodies (e.g., GM1 and Hu antibodies) in the workup of patients with chronic, indolent, sensory predominant, length-dependent polyneuropathies. CSF examination is usually normal and is also unwarranted.

In people with a large fiber neuropathy, the sensory NCS reveal either absent or reduced amplitudes that are worse in the legs.^1,3,4,6–12 Sensory NCV are normal or only mildly slow. Quantitative sensory testing (QST) demonstrates abnormal thermal and vibratory perception in as many as 85% of patients.^7,9 In addition, autonomic testing (e.g., quantitative sudomotor axon reflex and heart rate testing with deep breathing or Valsalva) is abnormal in some patients. Despite the fact that sensory symptoms predominate, motor NCS are often abnormal. Wolfe et al.⁹ reported that 60% of their patients with idiopathic polyneuropathy had abnormal motor NCS. The most common motor abnormalities are reduced peroneal and posterior tibialis compound muscle action potentials (CMAP) amplitudes, while distal latencies and conduction velocities of the peroneal and posterior tibial CMAPs are normal or only slightly impaired. Abnormalities of median and ulnar CMAPs are much less common. Fibrillation potentials and positive waves on needle EMG are also commonly found in intrinsic foot muscles as a further indicator of frequently subclinical motor involvement. In the authors’ experience, they may be the only indicator of motor involvement in what may otherwise appear to be a pure sensory neuropathy.

In patients with pure small fiber neuropathies, motor and sensory NCS are, by definition, normal. The peripheral autonomic nervous system is often affected in small fiber neuropathies; thus, autonomic testing can be useful.^13,24–27 The quantitative sudomotor axon reflex test (QSART) can be performed in the distal and proximal aspects of the legs and arms (Fig. 22-1). Sweat glands are innervated by small nerve fibers, and impaired QSART is highly specific and sensitive for small fiber damage, with 59–80% of patients having an abnormal study (Table 22-2).^24–27 Other autonomic tests [e.g., heart rate (HR) variability with deep breathing (DB) or Valsalva maneuver] may also be abnormal in affected individuals.⁷ In this regard, assessments include variability of HR to DB (Fig. 22-2) and response of the HR and blood pressure to Valsalva maneuvers and positional changes (e.g., response to tilt table or supine to standing position).

Abnormal thermal and vibratory perception thresholds may be demonstrated using QST.²⁸ Unlike NCS that only assess the physiology of large-diameter sensory fibers, QST of heat and cold perception can evaluate small fiber function. Abnormal QST has been reported in 60–85% of patients with predominantly painful sensory neuropathy (Table 22-2).^9,25,29,30 However, QST depends on patient attention and cooperation; it cannot differentiate between simulated sensory loss and sensory neuropathy. Furthermore, the sensitivity and specificity of QST are lower than QSART and skin biopsies.^31,32

HISTOPATHOLOGY

Nerve biopsies in patients with chronic, sensory predominant, length-dependent neuropathies may reveal axonal degeneration, regenerating axonal sprouts, or axonal atrophy with or without secondary demyelination.^5–7,9,33 Quantitative morphometry may reveal loss of large- and small-diameter myelinated fibers and small unmyelinated fibers. Occasionally, scattered perivascular and endoneurial lymphocytes may be seen on nerve biopsy,^33,34 although necrotizing vasculitis is not a feature. A clonal restriction of the variable T-cell receptor γ-chain gene has been demonstrated by one group of researchers.³⁵ Basal lamina area thickness, endoneurial cell area, and number of endothelial cell nuclei may be increased. However, the abnormalities on nerve biopsy are nonspecific and are generally not helpful in finding an etiology for the neuropathy. There, we do not routinely perform nerve biopsies on all patients with unexplained polyneuropathies. We consider doing a biopsy in people with autonomic sign or monoclonal gammopathies to assess for amyloidosis, those with multiple mononeuropathies, and in patients with underlying diseases associated with vasculitis (e.g., connective tissue disorders, cryoglobulinemia, and hepatitis B or C).

Nerve biopsies in individuals with small fiber neuropathies may show selective loss of small myelinated nerves and unmyelinated nerve fibers, but this requires quantitative analysis by electron microscopy (Fig. 22-3).¹³ A more sensitive and less invasive means of assessing these small fiber neuropathies histopathologically is by measuring intraepidermal nerve fiber (IENF) density on skin biopsies (Fig. 22-4).^{3,7,29,36–42} Assessment of IENF density also appears to be more sensitive in identifying patients with small fiber neuropathies than sural nerve biopsies, NCS, or QST (Table 22-2). Punch biopsy of the skin can be obtained at the foot, calf, or thigh, and immunohistochemistry using antibodies directed against protein gene product 9.5 (PGP 9.5) is used to stain small intraepidermal fibers. Intraepidermal nerve fibers arising entirely from the dorsal root ganglia represent the terminals of C and Aδ nociceptors. The density of these nerve fibers is reduced in patients with small fiber neuropathies, in which NCS, QST, and routine nerve biopsies are often normal. In at least a third of people with painful sensory neuropathies, IENF density on skin biopsies represents the only objective abnormality present following extensive evaluation.⁷