Clinical manifestations

Presentation and course

It is important to define polyneuropathy as a functional or structural disturbance of peripheral nerves causing motor, sensory, or autonomic symptoms and signs. This article pertains only to polyneuropathies that do not have an identifiable cause and are, thus, considered idiopathic. At least for idiopathic sensory polyneuropathies, standardized diagnostic criteria have been proposed (28).

Clinically, they typically present as length-dependent polyneuropathies with insidious onset of numbness or pain involving the feet. In this article, we divide idiopathic polyneuropathies into three clinical categories: sensorimotor polyneuropathy, pure sensory polyneuropathy (large fiber, small fiber, and mixed), and sensory neuronopathies, as the symptoms, exam findings, and diagnostic results differ between these groups. Routine use of a validated scoring system (Erasmus Polyneuropathy Symptoms Score) as a screening tool in clinical interviews is useful in identifying patients with a polyneuropathy in the general population (40).

In patients with sensorimotor polyneuropathies, pain is reported in 27% to 42% of patients, sensory loss in 65%, paresthesias in 33% to 68%, weakness in 26% to 82%, and difficulty with balance in 33% (62; 67; 49; 47; 60). Numbness is the most common presenting symptom. In a series of patients with sensory-only polyneuropathy or small fiber polyneuropathy, the most common symptoms are pain in 54% to 100% of patients, sensory loss in 86%, and paresthesias in 86% to 100% (67; 31; 74; 99; 47). In both groups, lower extremity symptoms usually precede upper extremity symptoms (67; 99). Patients who experience pain have higher sural amplitudes, and the pain seems to have a nonlinear relationship with the severity of the polyneuropathy (01). Autonomic symptoms are not prominent.

On physical examination of patients with idiopathic sensorimotor polyneuropathy, the most common physical examination findings are reduced or absent lower extremity vibratory sense in 80% to 100% of patients, pinprick in 53% to 95%, light touch in 59% to 91%, and proprioception in 9% to 18%. Sensory abnormalities in the hands occur in 36% of patients. On motor exam, there is distal lower extremity weakness in 39% to 100% of patients and hand weakness in 10%. Reduced or absent ankle jerk deep tendon reflexes is reported in 78% to 100% of patients (62; 67; 99; 49; 60). Distal atrophy may be present. Patients with loss of upper extremity reflexes typically have absent reflexes in the lower extremities. Differences in patient classification and inclusion criteria can explain much of the variability between series in reporting frequencies of symptoms and examination findings.

As opposed to the chronic, slowly progressive course of idiopathic sensory and sensorimotor polyneuropathies, idiopathic sensory neuronopathies may present as an acute to subacute process. Patients present with profound, often asymmetric, vibratory, and proprioceptive loss and with early sensory ataxia (16; 35). Deep tendon reflexes are absent, but muscle strength is preserved, although some patients will use the term "weakness" when describing their ataxia. Extraocular motility and pupillary reactivity may be affected in some patients. A prior review of patients with sensory neuronopathies noted the most common clinical findings to be ataxia in the lower or upper limbs, asymmetric distribution of sensory loss, and sensory loss not restricted to the lower limbs (11).

A sensory ganglionopathy involving predominantly small fiber function in a segmental fashion has also been described. Proximal clinical symptoms include facial pain and segmental sensory loss. A non-length-dependent pattern of fiber loss on skin biopsies assessing epidermal nerve fiber density may be seen (57; 30).

Prognosis and complications

Most patients with idiopathic sensory and sensorimotor polyneuropathy follow a stable to a slowly progressive course over years (62; 32; 67; 70; 31; 99; 49; 47; 85). A plateau phase of relative stability is particularly characteristic of pure sensory polyneuropathies (67). Independent ambulation is maintained in nearly all patients (76; 85). However, walking canes and ankle bracing may be required over time in the sensorimotor population (70). The need for assistive devices in patients with pure sensory presentations appears uncommon.

In contrast, idiopathic sensory neuronopathies are progressive, with many patients becoming wheelchair-bound over time (16; 84). Recovery is partly dictated by etiology but is poor in most idiopathic cases. This is likely the result of irreversible damage to sensory ganglionic neurons early in the course and Wallerian degeneration in the dorsal root ganglia and dorsal columns of the spinal cord during more chronic stages of the disease.

Clinical vignette

A 68-year-old man presented with a 7-year history of numbness and burning pain in both feet. He denied a family history of similar symptoms. He had no visceral autonomic symptoms. On examination, he had decreased sensation to pinprick distal to the ankles bilaterally. Vibration sense was absent at the toes and diminished at the ankles. Position sense was intact. Muscle strength was normal. Deep tendon reflexes were present except for absent responses at the ankles. Extensive laboratory testing, including CSF analysis, excluded identifiable causes. Electrophysiologic studies demonstrated an axonal sensorimotor polyneuropathy. A sural nerve biopsy revealed a severe reduction in large and small myelinated fibers. His foot burning responded favorably to pregabalin.

Biological basis

Etiology and pathogenesis

Over a hundred causes of polyneuropathy have been identified (39). No cause was found in 20% to 30% of axonal chronic polyneuropathies. Idiopathic polyneuropathy is the second most common polyneuropathy after diabetes. For idiopathic polyneuropathy, the chance of finding an etiology decreases with age (95).

Idiopathic sensory and sensorimotor polyneuropathies are almost exclusively axonal on electrophysiologic and histologic examinations.

In prior series, no more than 6% of patients demonstrated demyelinating features on nerve conduction studies (62; 67; 99). Nerve biopsy tissue generally confirms the axonal nature of these polyneuropathies. Loss of myelinated fibers, occasional regenerating cluster formations, and other axonal degeneration and regeneration features are the typical findings (62; 67; 99). Endoneurial vessels in sural nerve biopsies may show increased basal lamina area thickness, endoneurial cell area, and number of endothelial cell nuclei (92). In some cases, especially idiopathic small fiber subgroups, sural nerve biopsies are normal (31; 46; 74). In such cases, intraepidermal nerve fiber density measurements on punch skin biopsy can provide objective evidence of polyneuropathy. Although biopsy findings typically demonstrate evidence of axonal damage, a series reported eight patients with cryptogenic sensory polyneuropathy who had pathologic findings consistent with chronic myelinopathy (13). Atypical chronic inflammatory demyelinating polyneuropathy and vasculitic polyneuropathy should be considered and ruled out as an etiology of an idiopathic neuropathy in the appropriate clinical setting.

The ability to transform exogenous compounds is orchestrated by the genetic variability of detoxification enzymes, which may play a role in individual susceptibility to polyneuropathy. Glutathione S-transferases are responsible for a step in the detoxification process by catalyzing the conjugation of glutathione with a large variety of organic compounds. Glutathione S-transferases also protect tissues from reactive oxygen damage. The GSTT1 null polymorphism (impaired metabolism of toxic compounds) may lead to peripheral sensory and motor axonal damage (59).

Three biomarkers of subclinical inflammation, chemokines (CCL7, CXCL10) and an inflammation-related membrane protein (Delta/notch-like epidermal growth factor-related receptor or DNER), were discovered to partially mediate the association between anthropometric measures (overweight, obesity) and distal sensorimotor polyneuropathies (83). An increasing role of metabolic syndrome as a risk factor in peripheral neuropathy has been recognized globally (51). Findings of HLA associations also suggest that the immune system has a role in the pathogenesis of idiopathic axonal polyneuropathy (101).

Studies have demonstrated that serum from patients with idiopathic sensory neuronopathy inhibits dorsal root ganglia neurite growth and immunostains fixed cultured and cryostat rat dorsal root ganglia, providing evidence for an immune-mediated pathogenesis in this small group of patients (94). Outside of paraneoplastic and autoimmune (Sjogren syndrome, systemic lupus erythematosus, autoimmune hepatitis, and possibly celiac disease) mechanisms, infectious (human immunodeficiency virus, Epstein-Barr virus, varicella-zoster virus, human T cell lymphotropic virus-1) and toxic mechanisms (pyridoxine, cisplatin, carboplatin, oxaliplatin) need to be considered. Up to 50% of sensory neuronopathies remain idiopathic (35).

Epidemiology

Investigation of the true prevalence of polyneuropathy in large populations is difficult. Data based on a review of 29 population-based studies (39) resulted in a large variability in prevalence rates: 0.1% to 12.6% across all ages, with higher prevalence rates in the elderly of 1.9% to 30.9%. This variability likely stems from the methods used to establish the diagnosis, specifically the sensitivity of the screening procedure, geographical location of the survey, life expectancy, and year the study was performed. Overall estimated prevalence of polyneuropathy in the general population is 1% to 3% and increases to 7% in the elderly. Both socioeconomic status and age distribution influence prevalence rates. Men are more commonly affected by idiopathic polyneuropathy than women (93). The percentage of idiopathic polyneuropathy has decreased over time (102). Polyneuropathy-caused disease burden, even adjusting for comorbidities, will worsen as populations evolve to older demographics (43). Polyneuropathy, including idiopathic polyneuropathy, is a disabling condition and may independently contribute to increased mortality (43).

Chronic, acquired, sensory, and sensorimotor polyneuropathies are common in middle and late adulthood, with an estimated prevalence of more than 3% (08). The incidence of idiopathic polyneuropathy goes from 1 in 7 at age 40 to 49 to 1 in 3 at age 80 years and older (95). The majority of acquired polyneuropathies are secondary to readily identifiable causes, such as diabetes, alcohol abuse, and a wide variety of metabolic and toxic disorders (93), including neurotoxic medications. After known etiologies are excluded, a large fraction of polyneuropathies remain idiopathic. Contemporary series of patients evaluated for sensorimotor polyneuropathy estimate 10% to 25% to be idiopathic (22; 62; 67; 99; 78).

The percentage of patients that are with idiopathic cause appears to be higher in groups of sensory-only or small fiber neuropathy, with some series reporting up to 90% without known etiology (74). However, studies that included testing for impaired glucose tolerance and celiac disease found 50% to be idiopathic (18).

The onset of idiopathic polyneuropathies occurs mainly in the sixth and seventh decades of life. The mean age of patients in published series ranges from 51 to 63 years of age (62; 70; 74; 99; 47; 60; 18). There appears to be no difference in the age of onset between sensorimotor and sensory groups.

In a study of small fiber polyneuropathy, half of whom were idiopathic, there was a greater number of women than men (18).

Factors associated with idiopathic polyneuropathies may reflect underlying etiologies. In a study, normoglycemic patients with neuropathy had more dyslipidemia than diabetic patients without neuropathy (87). A number of studies support the contribution of hyperlipidemia and the metabolic syndrome to the risk of developing polyneuropathy (47; 91; 83). Population studies suggest an association of polyneuropathy with the metabolic syndrome, specifically waist circumference and elevated triglyceride levels (38). Interestingly, in the same study there was no association between impaired fasting glucose and polyneuropathy, but the former was associated with lower sural amplitudes. Schlesinger and colleagues studied the incidence of distal sensorimotor polyneuropathy in a population of 513 participants (83). After a mean follow-up of 6.5 years, 127 cases of distal sensorimotor polyneuropathy were detected. They found that general and abdominal obesity were associated with incident polyneuropathy among individuals with and without diabetes. Importantly, a diagnosis of polyneuropathy is independently associated with all-cause mortality in the general population, even in the absence of diabetes (42).

Estimates of the prevalence of idiopathic sensory neuronopathies are less well-defined. However, studies examining the etiology of sensory neuronopathies have found a significant proportion to have no known etiology. In a study of 78 patients with sensory neuronopathy, 22 patients (28%) were determined to be idiopathic (11), although other estimates get closer to 50% (84).

Prevention

It is important to educate patients about the potential risk of developing a sensorimotor polyneuropathy if they gain weight, have hyperlipidemia, or have glucose intolerance or diabetes. In the study by Schlesinger and colleagues, the odds ratios for developing a sensorimotor polyneuropathy were 3.06 for overweight and 3.47 for obesity (83). A 5 cm increment in waist circumference had a 1.4 odds ratio for developing polyneuropathy. It is, therefore, advisable to promote weight loss in order toprevent a sensorimotor polyneuropathy.

Differential diagnosis

Reevaluations of idiopathic polyneuropathy may eventually uncover a cause in a sizable number of patients. Dyck and colleagues found that intensive evaluation of referred unclassified polyneuropathies yielded an identifiable cause in 76% (22). Of 205 patients referred for evaluation, 42% had hereditary polyneuropathies, 21% had inflammatory-demyelinating polyradiculoneuropathies, and 13% had other acquired neuropathies. In another study, the diagnosis of idiopathic axonal polyneuropathy was revised in nearly one half of 53 patients followed for at least 3 years (78). Other studies, however, have shown reevaluation of idiopathic polyneuropathies to be low-yield. In a study of 75 patients with chronic axonal polyneuropathy followed for 5 years, only four patients (5%) had an identified cause of neuropathy (70). Another series of 40 patients followed for at least 4 years found no etiology for any patient (49). Differences in referral populations and the breadth of the initial workup are likely responsible for the contrasting yields of reevaluations between studies.

Impaired glucose tolerance is one of the more common and increasingly recognized etiologies of previously categorized idiopathic polyneuropathies. A number of studies have reported increased prevalence of impaired glucose tolerance in patients with sensory polyneuropathy, even in the absence of frank diabetes (23; 71; 86; 89; 64). A study comparing diabetic polyneuropathy with idiopathic polyneuropathy found that the former had greater involvement of pure small fibers, whereas the latter had greater involvement of pure large fibers and a higher frequency of pain phenomena (48).

Hereditary motor and sensory polyneuropathy or Charcot-Marie-Tooth disease type 2 should be considered when motor symptoms predominate and skeletal deformities are present. Muscle cramping in the legs and feet and the absence of paresthesias favor a hereditary polyneuropathy over other etiologies (22). Onset late in life does not exclude hereditary motor and sensory polyneuropathy, as it is estimated that 15% of patients with autosomal dominant hereditary motor and sensory polyneuropathy type 2 present after 50 years of age (67).

The differential diagnosis for idiopathic sensory neuronopathies also includes paraneoplastic neuronopathies associated with anti-Hu antibodies; immune-mediated, including chronic ataxic neuropathy with disialosyl antibodies or CANDA; pyridoxine- or chemotherapy-induced toxicity; vitamin E deficiency; acute nutritional axonal neuropathy; and Sjogren syndrome (82; 33; 17; 35; 37).

Confusing conditions

On clinical grounds, idiopathic sensory and sensorimotor polyneuropathies may be indistinguishable from polyneuropathies with an identifiable cause. Therefore, the clinician must exclude potential causes, including diabetes, chronic alcoholism, metabolic disturbances, endocrine abnormalities, connective tissue diseases, malignancy or amyloidosis, HIV or other infections, pertinent toxic or pharmacologic exposure, and hereditary factors. Celiac disease must also be considered, as it has been associated with a painful axonal sensory polyneuropathy even in the absence of gastrointestinal symptoms (14).

Diagnostic workup

Idiopathic polyneuropathies, by nature, are a diagnosis of exclusion, which is established after a careful medical, family, and social history, neurologic examination, and directed laboratory testing to rule out the many potential causes of a length-dependent neuropathic process (73).

Laboratory testing. Declaring a polyneuropathy idiopathic rests many times on performing a set of blood tests. The approach goes back to the recognition that peripheral nerve tissue is sensitive to various pathogenic factors arising from systemic disorders. Historically, the list of disorders causing a polyneuropathy has grown over time, explaining perhaps why this topic requires periodic reviews.

The basic and highest yield laboratory studies in distal sensory or sensorimotor polyneuropathy include fasting blood glucose, vitamin B12 level, copper level, and serum protein immunofixation electrophoresis. If those studies are normal, oral glucose tolerance testing should be completed, especially if the clinical presentation includes pain (24). Other laboratory studies also recommended in routine lab testing include a basic metabolic panel, complete blood count, erythrocyte sedimentation rate, antinuclear antigen, rheumatoid factor, and thyroid function testing. The following studies should also be considered in the appropriate clinical situation: serum levels for vitamin B1, B2, vitamin E; nicotinic acid serum protein immunofixation; urine protein immunofixation; antineutrophil cytoplasmic antigen (ANCA) antibodies; serologies for IgG and IgA antigliadin antibodies; IgA anti-transglutaminase antibodies; human immunodeficiency virus (HIV) antibody; Lyme antibodies and confirmatory western blot, if applicable; heavy metal levels; and paraneoplastic and other antibody assays (73).

Testing for vitamin B12 deficiency is relatively high-yield and should include testing levels of the cobalamin metabolites, methylmalonic acid, and homocysteine. In two series of patients with polyneuropathy, between 2.2% and 8% of patients had B12 deficiency (06; 80), and in 5% to 10% of patients with serum B12 levels in the low to normal range, methylmalonic acid and homocysteine levels were elevated (02; 81). Testing for B6 levels may be left to the clinician’s discretion, with the caveat that moderately elevated plasma levels (100-200 mcg/L) are not associated with worse polyneuropathy signs or symptoms (88).

The incidence of monoclonal proteins in chronic idiopathic polyneuropathy is as high as 10%, representing a monoclonal gammopathy of uncertain significance in most patients (53; 10; 52; 77). Patients with an IgM monoclonal gammopathy commonly have specific antibody binding against peripheral nerve antigens, implicating an autoantibody pathogenesis of associated neuropathies (54). Polyneuropathy associated with IgM paraprotein may have a distinct distal demyelinating phenotype (DADS-M). However, patients with axonal polyneuropathy patients with monoclonal gammopathy of uncertain significance are difficult to distinguish on clinical and electrophysiological grounds from patients with idiopathic polyneuropathy without a paraprotein (69), raising questions about the pathogenic role of the immunoglobulin and whether the monoclonal gammopathy of uncertain significance, particularly IgG and IgA, is a coincidental discovery. Incidental paraproteins are found in up to 3% of the general elderly population (53), the age group with the highest prevalence of polyneuropathies.

In earlier studies of idiopathic sensory-predominant polyneuropathy, approximately one quarter of patients selected from case records or a clinic population had anti-sulfatide antibodies (75; 66). However, other series have found anti-sulfatide antibodies in none (67; 99) or only a small percentage of the patients with idiopathic polyneuropathy (68; 74). Anti-sulfatide antibody testing does not have a clear associated phenotype or treatment and should not be pursued routinely in clinical practice at this time.

CSF analysis is largely unremarkable in idiopathic polyneuropathies when inflammatory causes are excluded. The mean CSF protein from 73 unclassified patients in a series was 43 mg/dl (67). CSF was normal in four of five patients from another series (99). In a 2001 series of eight patients who had lumbar puncture for idiopathic neuropathy, none had elevated protein (49). Although typically normal in idiopathic polyneuropathies, CSF analysis should only be considered in patients with progressive symptoms or if there is clinical suspicion of atypical chronic inflammatory demyelinating polyneuropathy.

Genetic testing for hereditary peripheral polyneuropathies, such as Charcot-Marie-Tooth disease, should also be considered in patients who have the classic phenotype of such syndromes. Studies have shown genetic mutations in 54% to 100% of patients with the clinical diagnosis of Charcot-Marie-Tooth disease (98; 65; 58; 15). There may be a genetic etiology to a peripheral neuropathy even in the absence of family history due to the presence of de novo mutations. Guidelines recommend electrodiagnostic studies to determine whether the polyneuropathy is axonal or demyelinating to guide testing for genetic mutations in the patient. The role of genetic testing in patients with idiopathic polyneuropathy without the classic phenotype is unknown (24a).

Nerve conduction studies and electromyography. Nerve conduction studies are indicated to define the phenotype, establish severity, and determine whether a polyneuropathy is axonal or demyelinating and focal or generalized. Nerve conduction studies in idiopathic sensory and sensorimotor polyneuropathies are almost exclusively axonal.

Nerve conduction abnormalities are expected in most patients with idiopathic polyneuropathy but may be normal in cases of small-fiber neuropathy. In prior series, at least 75% of patients with idiopathic polyneuropathy had abnormal nerve conduction studies (62; 67; 99). As expected, sural sensory responses are most frequently affected. Nerve conduction abnormalities typically consist of reduced amplitudes with normal or minimal distal latency and conduction velocity changes. The radial sensory and tibial motor action potential values seem to be the strongest electrophysiological determinant of polyneuropathy severity (100).

Most patients demonstrate abnormalities on needle electromyography, even if there is no evidence of motor nerve involvement by history or examination. In a large study, 70% of patients had an abnormal electromyography (99). Fibrillations were present in 42% and neurogenic motor unit potentials in 63%. Electromyography abnormalities have also been found in smaller series of patients with pure sensory presentations (31). Therefore, it is not uncommon for patients with idiopathic polyneuropathy who only have sensory symptoms and signs to demonstrate subclinical motor involvement on electrophysiologic studies.

Nerve conduction abnormalities are less common in patients with pure small fiber forms, with a yield of only 50% in patients presenting with painful, burning feet and minimal physical signs of neuropathy (31; 74).

Quantitative sensory testing. Quantitative sensory testing has been applied in several studies of idiopathic polyneuropathies and is the most commonly used approach for the functional assessment of afferent sensory fibers. By measuring thermal thresholds, quantitative sensory testing has a potential advantage over routine nerve conduction studies in assessing small sensory fibers, the fiber population predominantly affected by many idiopathic polyneuropathies. Quantitative sensory testing commonly demonstrates abnormalities in patients with neuropathy symptoms (45; 74), and it was slightly more sensitive than nerve conduction studies in one large study (99). However, another study observed no concordance between quantitative sensory testing and intraepidermal nerve fiber density measurements (74). Quantitative sensory testing also requires subjective patient input and may reflect abnormalities outside of the peripheral nervous system (34; 26; 27). There is substantial variation among quantitative sensory testing techniques and this, in turn, introduces another element of heterogeneity (05; 50).

Autonomic testing. Autonomic testing should be considered for patients with polyneuropathy, particularly if there is clinical evidence for autonomic polyneuropathy or distal small-fiber sensory polyneuropathy. Measurements of heart rate variability and quantitative sudomotor axon reflex testing (QSART) have been shown to be sensitive and specific for polyneuropathy, especially when used in combination (25).

Imaging. Spinal cord diffusion tensor imaging enables in vivo detection of posterior column damage in patients with sensory neuronopathy compared to normal subjects and distal diabetic polyneuropathy (12).

Nerve biopsy. Nerve biopsy generally confirms the axonal nature of idiopathic polyneuropathies. All 31 nerve specimens from a large series showed axonal degeneration (67). There was no evidence of demyelination, inflammation, vasculitis, or amyloidosis. Of 14 sural nerve biopsies from another large series, 13 demonstrated typical features of axonal degeneration (99). These studies suggest that sural nerve biopsy is rarely helpful in patients with idiopathic sensory neuropathies.

Nerve biopsy should be considered, however, when there is clinical suspicion for certain types of neuropathies, including amyloid neuropathy, mononeuritis multiplex due to vasculitis, or atypical chronic inflammatory demyelinating polyneuropathy (CIDP) (25). A series of eight patients was reported with a presentation of sensory neuropathy, predominantly large fiber distal sensory loss on examination, normal muscle strength, and minimally abnormal or axonal pathology findings on electrodiagnostic testing. These patients had sural nerve biopsy demonstrating chronic myelinopathy, consistent with a diagnosis of CIDP (13). Patients with progressive, predominantly large fiber idiopathic sensory neuropathy should be considered for sural nerve biopsy because CIDP is a treatable disease that often responds to therapy.

Skin biopsy. Skin biopsy for measurement of intraepidermal nerve fiber (IENF) density should be considered in the workup of distal sensory neuropathy, particularly in cases of small fiber sensory neuropathy. The most commonly used technique involves a 3 mm punch biopsy of skin from the leg, which is sectioned and examined using immunostaining techniques for anti-protein-gene-product 9.5 (PGP 9.5) antibodies to count the IENF density (24). Skin biopsy for IENF density has been shown to be reliable and valid in patients with distal sensory neuropathy, particularly in patients with small fiber neuropathy, with a reported diagnostic efficiency of 88% (61; 19). It is more sensitive than quantitative sensory testing or sudomotor autonomic testing, and one series demonstrated reduced IENF density in three quarters of patients with painful sensory symptoms and normal nerve conduction studies (74). In another study, IENF on skin biopsy was abnormal in 39% of 158 patients with clinically suspected small fiber neuropathy, of which 50% were of idiopathic etiology (18). High concordance has been reported between reduced IENF density and loss of pinprick on clinical examination (96). Skin biopsy for IENF density has also been shown to be more sensitive for small fiber sensory neuropathy than sural nerve biopsy, but it is typically normal in demyelinating neuropathies (41). Guidelines from both the American Academy of Neurology (AAN) and European Federation of Neurological Societies (EFNS) have recommended that skin biopsy for determination of IENF density is a valid and reliable method for diagnosis of distal sensory neuropathy, particularly small fiber sensory neuropathy (55; 24). Skin biopsy should be strongly considered in patients when such diagnoses are being considered. A significant heterogeneity is found in associations between fiber density, functional characteristics, and symptomatology. The severity of epidermal denervation is consistently associated with objective signs rather than pain or symptom scores (50).

The diagnostic workup of idiopathic sensory neuronopathies focuses on excluding identifiable causes. Laboratory testing should specifically include evaluation for vitamin B6 toxicity; HIV, HTLV, EBV, and VZV infection; Sjogren syndrome; systemic lupus erythematosus; paraneoplastic etiology; and GD1b antibodies. In patients with large-fiber sensory neuronopathy, cervical spine MRI often shows signal abnormalities in the dorsal columns and can be considered to support the clinical diagnosis (56; 57). Two studies examining IENF density in sensory ganglionopathies have shown reductions in the proximal thigh greater than or equal to that of the distal leg, highlighting the non-length-dependent pattern of the process (57; 30).

Corneal confocal microscopy. Recognition that corneal innervation is associated with epidermal nerve small fiber loss in the legs has prompted the use of corneal confocal microscopy as a less invasive method of diagnosis (50).

If clinically appropriate, mutation analysis for transthyretin familial amyloidosis should be pursued as this is now a treatable disorder (36).

Management

Given incomplete understanding of the pathogenesis of chronic idiopathic axonal polyneuropathy, there is no current disease-modifying treatment (97).

Supportive measures are discussed in Peripheral neuropathies: supportive measures and rehabilitation.

There is no current recommendation for dietary modifications for patients with chronic idiopathic axonal polyneuropathy (90).

Treatment of neuropathic pain is the primary focus of management for most patients with idiopathic sensory and sensorimotor polyneuropathies. Treatment studies for neuropathic pain have been conducted primarily in diabetic and HIV-associated neuropathies and postherpetic and trigeminal neuralgias.

A number of review articles have stratified medications for painful polyneuropathy into first-, second-, and third-line therapies based on the evidence supporting their benefit and potential risks and side effects (04; 21; 63; 72; 73). All agree that first-line therapies for peripheral neuropathic pain include tricyclic antidepressants and calcium channel alpha2-delta ligands (gabapentin and pregabalin). Topical lidocaine, available in patch and gel forms, is considered first- or second-line therapy for localized peripheral neuropathic pain, with the greatest evidence for postherpetic neuralgia. Antidepressants with dual serotonin and norepinephrine reuptake activity (duloxetine and venlafaxine) are first-line therapy in some reviews and second-line in others. Opioid medications and tramadol have been early on demonstrated to have good efficacy for neuropathic pain, particularly in combination with gabapentin (29), but most authors consider these agents to be second- or third-line therapies based on concern for side effects and dependence. However, retrospective data analysis suggests that the long-held practice of using chronic opioids for neuropathic pain to improve outcomes is not supported by facts. Long-term opioid use for neuropathic pain caused tolerance requiring escalating doses, hyperalgesia, addiction, and physical dependence resulting in behavior aimed at continuing opioids to prevent withdrawal. Chronic opioid use for neuropathic pain did not foster any functional impairment and caused increased adverse outcomes (44).

Other medications that are typically classified as third- or fourth-line therapies with inconsistent results for efficacy include other antidepressants (selective serotonin reuptake inhibitors), other antiepileptics (carbamazepine, oxcarbamazepine, lamotrigine, valproate, topiramate), topical capsaicin, cannabinoids, mexiletine, clonidine, and N-methyl-D-aspartate receptor antagonists (memantine, dextromethorphan). Agents demonstrating efficacy in trigeminal neuralgia have a separate body of evidence and recommendations regarding therapy.

A comparative effectiveness study of neuropathic pain treatment in patients with cryptogenic sensory polyneuropathy evaluated nortriptyline, duloxetine, pregabalin, or mexiletine. Even though there was no clearly superior medication, the highest utility was for nortriptyline followed in order by duloxetine and pregabalin. Of note is that the most efficacy was 25%, outweighed by quitting treatment rates of 37.3% to 58% (07). Mexiletine had the highest quit rate (58%) due to gastrointestinal side effects, but it had the highest probability (> 90%) of improvements in pain and fatigue if used to 12 weeks (09).

There are limited data to support the empiric use of immunosuppressive agents in patients with predominantly sensory idiopathic polyneuropathies. There have been individual reports of improvement of pain or motor function with intravenous immune globulin (31; 20) or other immunosuppressants, but evidence is limited and empiric treatment without known etiology is generally not recommended, particularly if the course is benign.

Patient education plays an important role in the management of idiopathic polyneuropathy. Patients are often anxious about future loss of function. The clinician can be a source of reassurance by referring to available literature stating that the vast majority of patients remain stable or progress slowly over time, suffer limited motor disability, and have a relatively favorable long-term prognosis. For those patients with pure sensory polyneuropathies, the prognosis may be even more favorable with long plateau periods. Simply relaying this natural history to patients can provide considerable emotional and even physical comfort. Emphasis on weight loss, treatment of hyperlipemia, and daily exercise to promote weight loss and improve strength and balance should also help.

For patients with classified or idiopathic sensory neuronopathies, therapy depends on presumed etiology (36). For idiopathic cases corticosteroids, oral immunosuppression, IVIG, and plasma exchange have all been reported from small series, and no benefit was recorded; however, most patients were treated late in their course (35). Treatment of neuropathic pain remains a “trial-and-error” approach as phenotypic classification does not fully predict response to pharmacological interventions (03).

Outcomes

Given the incomplete understanding of the pathogenesis of chronic idiopathic axonal polyneuropathy, there is no current disease-modifying treatment (97).