Article Text

Download PDFPDF
Untangling a case of painful neuropathy
  1. Stephen Keddie1,2,
  2. Janev Fehmi3,4,
  3. Zane Jaunmuktane5,
  4. Shirley D'Sa6,
  5. James C Stevens4,
  6. Michael PT Lunn1,2
  1. 1 National Hospital for Neurology and Neurosurgery, MRC Centre for Neuromuscular Diseases, London, UK
  2. 2 National Hospital for Neurology and Neurosurgery, Institute of Neurology Department of Neuroimmunology and CSF Laboratory, London, UK
  3. 3 Nuffield Department of Clinical Neurosciences, John Radcliffe Hospital, Oxford, UK
  4. 4 Department of Neurology, North Bristol NHS Trust, Westbury on Trym, Bristol, UK
  5. 5 Department of Neuropathology, National Hospital for Neurology and Neurosurgery, London, UK
  6. 6 Department of Haematology, Cancer Division, University College London Hospitals NHS Foundation Trust, London, UK
  1. Correspondence to Dr Michael PT Lunn, National Hospital for Neurology and Neurosurgery, MRC Centre for Neuromuscular Diseases, London, Greater London WC1N 3BG, UK; michaellunn{at}

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

A 51-year-old right-handed man developed a sensory neuropathy with the sensation of walking on pebbles, and tight cramping pains in the calves requiring amitriptyline. Two months later, he became unsteady with difficulty walking downstairs necessitating a stick, with associated fatigue and weight loss of half a stone. He had a history of haemophilia type A and liver cirrhosis from hepatitis C, contracted through the transfusion of blood products in the 1980s.

On examination, there was evidence of a distal symmetrical sensorimotor polyneuropathy with Medical Research Council (MRC) power grade 4+/5 in the intrinsic hand muscles and 0/5 in ankle plantar and dorsiflexion. Supinator, knee and ankle deep tendon reflexes were all absent, with mute plantar responses. He had a loss of pinprick sensation to the knees, vibration loss to the costal margins and temperature perception was altered to the mid-thighs. There was dusky pigmentation of the skin and peripheral oedema in the lower limbs.

What is the differential diagnosis, and which investigations should be considered at this stage?

His history and examination findings are typical of a distal symmetrical neuropathy (see box 1 for the differential diagnosis of a painful large fibre neuropathy). Serological workup for a neuropathy should include full blood count, B12, folate, erythrocyte sedimentation rate (ESR), glucose, thyroid function, a serum protein electrophoresis and immunofixation, and possibly an antinuclear antibody. Note that it is typical for hospital laboratories only to perform serum immunofixation if the serum protein electrophoresis is abnormal. The greater sensitivity of immunofixation may find low-level paraproteins that would be otherwise undetectable, and which may be highly relevant in the context of neuropathy. He also has a history of hepatitis C which is associated with cryoglobulinaemia and cryoglobulinaemic peripheral nerve vasculitis, and therefore an important differential. Constitutional symptoms, including weight loss, are uncommon in isolated peripheral neuropathy, and therefore systemic causes should be considered, including vasculitis, malignancies, metabolic (diabetes), infective and amyloidosis. Thus, a vasculitic screen, lactate dehydrogenase, HIV test, antineuronal and antiganglioside antibodies may also be useful.

Box 1

Differential diagnosis of a large fibre painful peripheral neuropathy

Diabetes (including lumbosacral plexopathy)

Alcohol overuse


  • Guillain-Barré syndrome (GBS)

  • Vasculitis

  • Some chronic inflammatory demyelinating polyradiculoneuropathy

  • Paraprotein-associated neuropathy

  • Polyneuropathy, organomegaly, endocrinopathy, M-protein and skin changes syndrome

  • Cryoglobulinaemic neuropathy

  • Idiopathic plexopathies


  • Hepatitis (B or C)

  • Cytomegalovirus

  • Varicella-zoster virus

  • HIV

  • Lyme disease

Mechanical/compressive neuropathies

Nutritional (B6 deficiency)

Metabolic (eg, Fabry disease)


  • Drugs (chemotherapies/antibiotics)

  • Arsenic

  • Mercury

  • Cadmium


  • Direct: local infiltration by prostate, melanoma and lung

  • Indirect: neurolymphomatosis


  • Hereditary transthyretin amyloidosis (hTTR)


  • P0

  • Rab7

Nerve conduction and electromyography are necessary to characterise the distribution and pattern of motor and sensory involvement, including differentiating demyelinating from axonal pathologies.

Neurophysiology identified a symmetrical length-dependent sensorimotor neuropathy with slowed conduction velocity in the upper limb, delayed F waves and axonal loss in the lower limbs, and denervation (table 1). Slowed conduction velocity and prolonged F-waves suggest a demyelinating disorder, of which chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is the most common, but this remains a diagnosis of exclusion. Other causes include genetic, drugs (amiodarone, and some immunomodulatory, immunosuppressive and antineoplastic agents), infective and paraprotein-associated neuropathies. This degree of axonal loss would be very atypical for CIDP at an early stage (unless in nodoantibody mediated-paranodopathies with reversible conduction failure), and given that it can result in conduction slowing as a secondary phenomenon from preferential large fibre loss, one should seek clarification from neurophysiology colleagues as to whether the conduction slowing recorded is explained by the degree of axonal loss, or representative of ‘true’ slowing.

Table 1

Neurophysiology and electromyography results

Blood tests identified type I cryoglobulinaemia. What is the significance of this result and the likelihood of causing neuropathy?

Cryoglobulinaemia is the presence of serum immunoglobulins that precipitate in vitro at temperatures below 37°C and re-dissolve after warming. The cryoprecipitate type is relevant to the diagnosis and is determined through immunofixation of the re-dissolved cryoprecipitate. Type I consists of monoclonal immunoglobulins, either IgM or IgG, can be associated with hyperviscosity syndrome and can cause vascular occlusion. Type II and III are ‘mixed’, with a combination of monoclonal IgM and polyclonal IgG, or polyclonal IgM and IgG. Cryoglobulins are pathogenic through the formation of immune complexes that bind to receptors on epithelial cells causing vascular inflammation. The leading cause of mixed cryoglobulinaemia is chronic hepatitis C infection but other infections such as HIV and hepatitis B, and autoimmune diseases, such as Sjögren’s syndrome, systemic lupus erythematosus and rheumatoid arthritis, are also associated. Multisystem clinical manifestations result from cryoglobulin precipitation and vascular inflammation in target organs, that is, skin (ulcers, gangrene, livedo reticularis, purpuric rash and others), vasomotor (Raynaud’s phenomenon and acrocyanosis), peripheral or central nervous system, musculoskeletal (arthralgias and arthritis) and renal (renal failure and proteinuria).1

Neurological involvement occurs in approximately 30% of patients with type I cryoglobulins, with 20%–60% of patients presenting with peripheral neuropathy as the initial symptom.1 Presentation is predominantly with a sensory neuropathy affecting both large and small fibres, resulting in painful paraesthesia, most commonly in the lower limbs, which can be asymmetric and is typically worse at night. Less commonly, patients present with multiple mononeuropathies, typified by a painful asymmetric, sequential peripheral nerve impairment. Motor symptoms do occur but often later.

This patient’s cryoglobulin immunofixation demonstrated type I cryoglobulins, which are monoclonal, and therefore typically driven by a B-cell haematological malignancy, such as B-cell non-Hodgkin's or Hodgkin's lymphomas, multiple myeloma, or chronic lymphocytic lymphoma. An IgG lambda paraprotein was discovered. Paraproteins are present in approximately 10% of patients with neuropathy with no other cause, though the presence of a demyelinating neuropathy with any paraprotein should warrant further investigation to determine its origin and potential pathogenic role.

What investigations are necessary following the identification of a monoclonal gammopathy?

A monoclonal gammopathy should be investigated to delineate its origin as monoclonal gammopathy of undetermined significance (MGUS), lymphoma, the lymphocytic leukaemias, multiple myeloma or light chain amyloidosis. MGUS is of considerable clinical importance at diagnosis due to an 18% risk of progression to myeloma or other plasma or lymphoid malignancy in 20 years, and its known causal association with several serious non-malignant disorders, including anti-myelin-associated glycoprotein (anti-MAG) neuropathy, other paraprotein-associated neuropathy, POEMS syndrome (polyneuropathy, organomegaly, endocrinopathy, M-protein and skin changes) and CANOMAD syndrome (chronic ataxic neuropathy, ophthalmoplegia, IgM paraprotein, cold agglutinins and disialosyl antibodies).

Isolated, low-risk MGUS patients do not need bone marrow biopsy or skeletal imaging. The Mayo Clinic risk stratification model defines low risk as serum M-protein<15 g/L, IgG isotype and normal serum free light chain ratio. In such patients, the probability of finding disease necessitating haematological treatment was 4.7%.2 The case presented here can be considered as a ‘monoclonal gammopathy of neurological significance’ in which the MGUS may be contributing to cause neuropathy (see box 2 for differentials). An MGUS that does not fit the clinical or immunological phenotypes of the diagnoses listed in box 2 is likely to be coincidental to the neuropathy than causative. Anti-MAG antibodies (in IgM cases), antiganglioside antibodies (including anti-disialosyl in CANOMAD syndrome), skeletal imaging and bone marrow biopsy would be necessary investigations. Skeletal imaging can be performed by X-ray and CT scan, or now whole-body MR scan, but fluorodeoxyglucose (FDG)-PET CT or MR scanning may be of more clinical utility, and can identify bone lesions, plasmacytoma and peripheral nerve neurolymphomatosis. Serum vascular endothelial growth factor (VEGF) is a proinflammatory cytokine often elevated in POEMS syndrome, and therefore an important diagnostic biomarker.

Box 2

The differential diagnosis of ‘monoclonal gammopathy of neurological significance’

  • Paraprotein-associated neuropathy causing distal acquired demyelinating neuropathy with or without anti-myelin-associated glycoprotein antibodies (IgM)

  • Paraprotein-associated neuropathy with chronic inflammatory demyelinating polyradiculoneuropathy phenotype (more likely IgA or IgG)

  • Polyneuropathy, organomegaly, endocrinopathy, M-protein and skin changes (POEMS) syndrome (IgA or IgG lambda light chain restricted)

  • Chronic ataxic neuropathy, ophthalmoplegia, IgM paraprotein, cold agglutinins and disialosyl antibodies syndrome (CANOMAD) (chronic sensory ataxic neuropathy with anti-disialosyl antibodies) (IgM)

  • Neurolymphomatosis (IgG, IgA or IgM)

  • Light chain amyloidosis (typically lambda)

  • Coincidental finding to the neuropathy

A skeletal survey and MR scan of the whole spine showed sclerotic sacral and pelvic bone lesions thought to represent bony metastases or myeloma. Bone marrow examination found only 3% plasma cells, which were occasionally binucleated and therefore atypical and likely clonal. Bone lesion biopsy was consistent with a plasmacytoma.

Serum VEGF was 850 pg/mL (upper limit of normal 771 pg/mL).

What is the significance of the VEGF result?

Serum VEGF is often in the several thousands in patients with POEMS. Serum concentrations above 771 pg/mL and less than 1000 pg/mL are more difficult to interpret. It is an extremely sensitive diagnostic test for POEMS syndrome, with a sensitivity of almost 100% and specificity of 92% in patients with demyelinating neuropathy and lambda light chain-associated paraproteinaemia, when >1000pg/mL but was only mildly elevated in this case. Given that the serum VEGF can be raised in cases of vasculitic neuropathy and other clonal haematological disorders, this result is not diagnostic. Other false positive causes of raised serum VEGF are iron-deficiency anaemia and chronic hypoxia secondary to chronic obstructive pulmonary disease or obstructive sleep apnoea.

At this stage, the patient has a clonal plasma cell proliferative disorder, demonstrated both in the definite plasmacytoma and likely in the bone marrow, combined with a painful sensorimotor neuropathy. From the investigations, it remained unclear whether his neuropathy was a result of cryoglobulinaemic peripheral nerve vasculitis, or secondary to POEMS syndrome.

Which test would be useful to rule in/out cryoglobulinaemic vasculitis?

A sural nerve biopsy showed depletion of large myelinated fibres, widespread active axonal degeneration and occasional regeneration without histopathological evidence of cryoglobulinaemic vasculitis, although given the patchy nature of vasculitis, it may be missed. Electron microscopy showed evidence of uncompacted myelin lamellae, which suggests but is not specific for POEMS syndrome (figure 1). Although the cryoglobulinaemia was a useful clue towards the presence of a haematological malignancy, the cryoglobulins were not deemed to be the cause of neuropathy.

Figure 1

Sural nerve biopsy. (A) H&E stained section shows widespread endoneurial oedema (green arrow). (B) Semithin resin section, stained with MBA-BF (methylene blue azure-basic fuchsin), shows a moderate reduction in density of large myelinated fibres and accentuates increased actively degenerating axonal profiles (yellow arrows). (C) Immunostaining with macrophage marker CD68 further highlights the myelin ovoids of degenerating axons. (D) Scanty perivascular lymphocytes are evident in the epineurium on immunostaining with UCHL1 (ubiquitin carboxyl-terminal esterase-1), but with no signs of vasculitis. (E and F) Ultrastructural examination shows uncompacted myelin in some of the large calibre fibres, which has a different appearance to widely spaced myelin (G and H) seen in anti-MAG neuropathies (myelin pathology highlighted with blue arrows). Uncompacted myelin lamellae are typified by separation of the major dense line, compared with widely spaced myelin; two or more wraps of myelin with a regularly separated intraperiod line and an intact major dense line. (I) Scale bar: 80 µm in (A), (C) and D; 20 µm in (B); 5 µm in (E) and (G); and 1 µm in (F) and (H). *Bottom diagram was taken from Bilbao J, Schmidt RE. ‘Biopsy diagnosis of peripheral neuropathy’. Consent to use image was provided by Robert Schmidt. Anti-MAG, anti-myelin-associated glycoprotein.

Thus, we diagnosed POEMS syndrome on the basis of the typical clinical and neurophysiological findings, an IgG lambda paraprotein, with associated skin changes (acrocyanosis—dusky red lower limbs), peripheral oedema, sclerotic bone lesions and mildly raised serum VEGF. He was treated with a melphalan-200 autologous stem cell bone marrow transplant. Subsequently, the paraprotein disappeared and serum VEGF normalised, in keeping with a complete haematological response. The neuropathic pain resolved, and he is now walking unaided.


POEMS syndrome is a rare condition typified by the presence of an inflammatory demyelinating neuropathy and a monoclonal plasma cell disorder, which is lambda light chain restricted. Pain in the lower limbs is a common but under-reported early clinical feature, typically described as a cramping sensation in the calves bilaterally, often mistaken for peripheral vascular disease, with peripheral nerve vasculitis as another possible differential. Although a rare association of an already rare disease, type I cryoglobulinaemic neuropathy has been previously described in POEMS syndrome.3

Very often, if a positive serum VEGF is discovered in the correct clinical context (typical neuropathy, lambda light chain paraprotein with or without other POEMS syndrome features), POEMS syndrome is highly likely, and nerve biopsy would not be necessary to investigate the cause once typical bone marrow or plasmacytoma histology demonstrates an abnormal plasma cell clone. However, in this case, a nerve biopsy was crucial to ascertain the cause of neuropathy where the differential diagnosis was of POEMS syndrome versus cryoglobulinaemic neuropathy of an alternative cause.

Nerve biopsy in POEMS syndrome displays loss of large and small myelinated fibre density. Electron microscopy showing uncompacted myelin lamellae suggests POEMS syndrome, present in 60%–80% of POEMS syndrome nerve biopsies. There were no features suggestive of vasculitis, and thus, the histopathological features of neuropathy were of POEMS syndrome rather than cryoglobulinaemia.

Key points

  • Severe axonal loss on neurophysiology is atypical in early chronic inflammatory demyelinating polyradiculoneuropathy (unless associated with nodo-paranodopathies) and alternative diagnoses should be considered.

  • Serum immunofixation is more sensitive to protein electrophoresis and may uncover a small but potentially neurologically significant paraprotein.

  • Anti-myelin-associated glycoprotein antibodies (in IgM cases), antiganglioside antibodies (including anti-disialosyl in chronic ataxic neuropathy, ophthalmoplegia, IgM paraprotein, cold agglutinins and disialosyl antibodies syndrome (CANOMAD), serum vascular endothelial growth factor (VEGF), skeletal imaging and bone marrow biopsy would be necessary investigations for a potential monoclonal gammopathy of neurological significance.

  • Serum VEGF is a highly sensitive and specific biomarker for polyneuropathy, organomegaly, endocrinopathy, M-protein and skin changes (POEMS) syndrome but can be mildly raised in vasculitic neuropathies, iron-deficient anaemia and chronic hypoxia.



  • Funding SK is funded by the Guarantors of Brain and the Association of British Neurologists. MPTL is supported by the National Institute for Health Research, University College London Hospitals, Biomedical Research Centre.

  • Competing interests None declared.

  • Patient consent for publication Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed by Rob Hadden, London, UK.

Other content recommended for you