Article Text

Download PDFPDF
CIDP presenting as recurrent severe back pain without weakness or sensory loss
  1. Dirk Bäumer1,
  2. David James Grant2,
  3. Ravi Knight3,
  4. Camilla Buckley4,
  5. David Bennett4,
  6. Simon Rinaldi4
  1. 1Wessex Neurological Centre, University Hospital Southampton, Southampton, UK
  2. 2Department of Neuroradiology, John Radcliffe Hospital, Oxford, UK
  3. 3Department of Clinical Neurophysiology, John Radcliffe Hospital, Oxford, UK
  4. 4Nuffield Department of Clinical Neurosciences, John Radcliffe Hospital, Oxford, UK
  1. Correspondence to Dr Simon Rinaldi, Nuffield Department of Clinical Neurosciences, John Radcliffe Hospital, Headley Way, Oxford OX3 9DU, UK; simon.rinaldi{at}


A 71-year-old woman presented with severe back pain, limb weakness and cranial nerve dysfunction associated with high cerebrospinal fluid (CSF) protein; we diagnosed Guillain-Barré syndrome and her symptoms completely resolved after intravenous immunoglobulin. Over the next 4 years, she had three further episodes of excruciating back pain accompanied by raised CSF protein, but without weakness, sensory loss, or abnormalities in routine nerve conduction studies. Sensory evoked potentials suggested proximal demyelination and lumbosacral plexus imaging suggested inflammation. We argue that this is a relapsing proximal polyradiculoneuropathy on the spectrum of chronic inflammatory demyelinating polyradiculoneuropathy.


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.


Chronic inflammatory demyelinating polyradiculoneuropathy (CIDP) is an immune-mediated peripheral nerve disorder with a prevalence of 2.8 per 100 000.1 ,2 In addition to the well recognised phenotype presenting with symmetrical, proximal and distal sensorimotor dysfunction, the CIDP spectrum probably includes several less common syndromes with asymmetrical or focal presentations, as well as pure motor or pure sensory presentations.3 ,4 Pure sensory presentations of CIDP can occasionally have pathology restricted to the sensory roots (chronic immune sensory polyradiculopathy), leading to a syndrome of large fibre sensory loss and sensory ataxia with normal peripheral nerve conduction but delayed somatosensory evoked potentials and nerve root enlargement on MRI.5 ,6

We present a patient with pre-existing ocular myasthenia who developed recurrent episodes of severe back pain, without weakness or sensory loss secondary to a relapsing inflammatory polyradiculopathy. This also appears to be part of the CIDP spectrum.

Case report

At the age of 35 years, a British Caucasian woman developed variable ptosis and diplopia, without bulbar dysfunction or limb weakness. She was diagnosed with ocular myasthenia and her edrophonium test was positive. Her symptoms were completely controlled with pyridostigmine and low-dose prednisolone (5 mg on alternate days).

When aged 71 years, she developed an upper respiratory tract infection needing antibiotics. Four weeks later, she developed severe excruciating lower back pain and burning discomfort in her thighs, followed within days by new diplopia, ptosis, facial weakness, mild proximal limb weakness and lower limb areflexia. Investigations showed serum sodium of 128 mmol/L while clinically euvolaemic, acellular cerebrospinal fluid (CSF) with a protein of 2.2 g/L, and mildly prolonged lower limb F waves (minimal F wave latency of the common peroneal nerve from extensor digitorum brevis 66 ms) with no other nerve conduction abnormality. We diagnosed Guillain-Barré syndrome, and she improved with intravenous immunoglobulin treatment (2 g/kg body weight), returning to her baseline after 2 months.

One year later, when aged 72 years, she presented with a 7-day history of severe lower back pain, radiating into both legs, with no numbness, tingling or burning. She felt unsteady but could walk independently. Strength and sensory examinations were normal, and her deep tendon reflexes were present. Her CSF was acellular with a protein of 2.7 g/L with matched oligoclonal bands. Her peripheral nerve conduction was normal. We suspected a Guillain-Barré syndrome relapse and gave intravenous immunoglobulin, and she fully recovered within 2 months. Repeat CSF 6 months later, while she was asymptomatic, was acellular with normal protein.

A further year later, she presented again with a 2-week history of progressive severe neck pain radiating into both arms. The arm pain had a deep, achy character, with no associated sensory loss or positive sensory phenomena. She felt unsteady but walked normally. There was no weakness or sensory loss, and her deep tendon reflexes were brisk and symmetrical with flexor plantar responses. Neuroimaging showed mild cerebral small vessel disease and cervical spondylosis with bilateral, right more than left C5, C6 and C7 foraminal stenoses, not justifying surgical intervention. Her CSF protein was again raised at 1.1 g/L. Her symptoms subsided within 2 months.

One year later she developed a dental infection needing antibiotic treatment and a tooth extraction. One week later, she experienced severe neck and bilateral shoulder pain, followed several days later by severe lower back and bilateral buttock pain. Her pain worsened on movement but was also severe at night. There were no other sensory symptoms. As before she felt unsteady, but could still walk and had no falls. Examination showed normal power and reflexes. Sensory examination was normal for light touch, pinprick and vibration sense. She made a few errors with great toe movements on testing joint position sense. Romberg's test was normal.

At this point, 4 years from the initial Guillain-Barré syndrome-like presentation, we performed some further investigations. CSF examination showed a protein of 2.3 g/L with no cells and matched oligoclonal bands. Nerve conduction studies were normal apart from mildly prolonged F waves (table 1). Somatosensory evoked potentials from the upper limbs were in keeping with a delay between Erb's point and the lower cervical spine, consistent with demyelination of proximal nerve segments (figure 1). Right-sided tibial nerve evoked potentials were well defined with prolonged central latencies. There was no intermediate lumbar potential; this result is less specific and non-localising, but also supports proximal demyelination. Lumbosacral plexus imaging showed thickened and enhancing extraforaminal nerve roots (figure 2). Serum anti-nuclear antibody, anti-neutrophil cytoplasmic antibody, complement, serum electrophoresis, immunofixation, free light chains, ganglioside antibodies and an endocrine screen were normal or negative.

Table 1

Nerve conduction studies

Figure 1

Somatosensory evoked potentials. Trace and latency data from right median study. Somatosensory evoked potentials show normal latency to Erb's point (N9 waveform), with delayed conduction between Erb's point and the lower cervical spine (N13 waveform), supporting proximal peripheral demyelination.

Figure 2

MRI of lumbosacral plexus. Sequential coronal images from a postgadolinium fat-saturated T1-weighted MRI of the lumbar and sacral spine and plexus. There is thickening and enhancement of the extraforaminal segments of the left L3, bilateral L4 and bilateral L5 nerve roots (red arrows). These findings are consistent with a polyradicular inflammatory process. There is no thickening within the lumbar canal itself, and no MRI evidence of neural compression at any level.

We again gave intravenous immunoglobulin and her symptoms settled as before.

Nerve conduction studies were normal, apart from mildly prolonged F waves from the right median and tibial nerves.

In summary, this patient presented with four episodes of severe back pain associated with raised CSF protein, each responding to intravenous immunoglobulin. While the first episode was reasonably diagnosed as Guillain-Barré syndrome, subsequent episodes had no accompanying weakness, sensory loss or reflex changes. Peripheral nerve conduction studies were essentially normal throughout, but somatosensory evoked potentials and imaging gave clear evidence for involvement of the proximal peripheral nervous system.


This patient developed a relapsing, immune-responsive, proximal, demyelinating polyradiculopathy with inflammatory but acellular CSF; we argue that it is best defined as CIDP. Current diagnostic criteria for CIDP incorporate atypical forms, and specifically pure sensory variants (including chronic immune sensory polyradiculopathy affecting the central process of the primary sensory neurone). In contrast to the 2005 edition, the 2010 revision of the European Federation of Neurological Societies/Peripheral Nerve Society CIDP guidelines7 also include delayed somatosensory evoked potentials as a supportive diagnostic criterion.8 MRI evidence of enhancing or enlarged nerve roots is also a supportive criterion, despite ongoing debate as to what constitutes enlargement and whether nerve root diameters in CIDP are significantly different to controls.9 Nevertheless, our case meets most of the five published supportive diagnostic criteria (table 2).

Table 2

Supportive diagnostic criteria for chronic inflammatory demyelinating polyradiculoneuropathy

We argue that the tibial and peroneal F wave latencies achieve the electrodiagnostic threshold for possible CIDP on two occasions, and with inclusion of the supportive criteria diagnostic confidence is elevated to ‘definite’. Nevertheless, there are some unusual features. The first episode was Guillain-Barré syndrome-like in onset. This is well described, but occurs only in a minority of CIDP cases.10–12 Acute onset CIDP is more likely than recurrent Guillain-Barré syndrome when the first deterioration happens 8 weeks or more from disease onset, when there are three or more deteriorations, and when independent walking is not lost, as in this case. Cranial nerve involvement favours Guillain-Barré syndrome, although this can occur in CIDP. Two episodes also had evidence of preceding infection, which again occurs in CIDP13 but more commonly in Guillain-Barré syndrome. The matched oligoclonal bands may reflect this systemic infection. The most striking feature was the severe pain during all episodes, once triggering an emergency neurosurgical referral for suspected acute cervical disc herniation. Pain is generally a red flag in CIDP14 and should prompt the clinician to consider alternative pathologies, such as vasculitis or POEMS (polyneuropathy, organomegaly, endocrinopathy, monoclonal protein, skin changes) syndrome. However, the prevalence of pain in CIDP may be underestimated15 and may be the presenting symptom of patients otherwise meeting standard criteria for CIDP. Misra et al described three patients with acute onset arm pain and weakness mimicking brachial neuritis, but with multifocal conduction block and prolonged F waves on nerve conduction studies.16 Boukhris et al17 describe five patients with CIDP meeting conventional diagnostic criteria whose first symptom was limb pain. Headache occurred in a case of CIDP with markedly raised CSF protein and papilloedema,18 as well as in a case of relapsing cranial neuropathy thought to be due to CIDP.19

Finally, our patient was initially diagnosed with ocular myasthenia gravis. We considered whether this could have been a fluctuating oculomotor cranial neuropathy. However, her eye movement disorder never conformed to a single cranial nerve dysfunction, cranial imaging showed no thickened cranial nerves, there was a documented response to edrophonium, and serum acetylcholine receptor antibodies, when available, were raised, even if just above the upper limit of normal. Myasthenia gravis and CIDP are both immune-mediated, but while myasthenia gravis is antibody mediated, CIDP involves complex cellular and humoral factors.20 It is therefore perhaps surprising that the co-occurrence of both conditions is not more common, with only occasional case reports.21 ,22

In summary, this patient had a relapsing polyradiculitis with minimal neurological features but prominent pain. This expands the phenotype of CIDP and highlights the role of both somatosensory evoked potentials and neuroimaging in investigating peripheral nerve disorders.

Key points

  • Chronic inflammatory demyelinating polyradiculoneuropathy can present with significant pain, although the clinician should consider other conditions.

  • Normal routine nerve conduction studies do not exclude a peripheral nerve disorder.

  • Neuroimaging and somatosensory evoked potentials help to investigate the proximal segments of the peripheral nervous system.



  • Contributors DiB wrote the paper. DJG created the MRI figure, reviewed and edited the manuscript. RK created the neurophysiology figure, reviewed and edited the manuscript. CB supervised the patient's care and edited the manuscript. DaB conceived the paper and edited the manuscript. SR conceived the paper and edited the manuscript and figures.

  • Competing interests None declared.

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed. This paper was reviewed by Robert Hadden, London, UK.

Linked Articles

  • Editors' commentary
    Phil E M Smith Geraint N Fuller

Other content recommended for you