← Library

Chronic Inflammatory Demyelinating Polyradiculoneuropathy

Registrar reference · EAN/PNS 2021 · the chronic sibling of GBS
Compiled Jul 2026
Verify doses & PBS
Steroids work here
Decision support only — not a substitute for neurology or the EAN/PNS guideline. CIDP is the treatable chronic cousin of GBS, and two errors dominate: giving up on steroids (they work here, unlike in GBS) and overdiagnosing it (a large share of "CIDP" is actually paraprotein neuropathy, a hereditary neuropathy, or a nodopathy that needs rituximab). Screen every patient for a paraprotein, and reconsider the label whenever treatment doesn't work. Verify all doses.
1 CIDP vs GBS — same family, different rules
FeatureGuillain-Barré (acute)CIDP (chronic)
TempoNadir <4 weeksProgresses or relapses over >8 weeks (4–8 = subacute)
CourseMonophasicChronic — progressive, stepwise, or relapsing
SteroidsIneffectiveEffective — first-line (the key flip)
Resp / autonomicProminent, dangerousUsually absent; patients often stay ambulant
TreatmentOne course of IVIG or PLEXInduction + ongoing maintenance

The overlap: acute-onset CIDP

2 Diagnose it — and don't overdiagnose it

The clinical picture (EAN/PNS 2021)

  • CoreProgressive/stepwise/relapsing symmetric proximal AND distal weakness with sensory involvement of ≥2 limbs, over ≥8 weeks; areflexia/hyporeflexia in all limbs; cranial nerves may be involved.
  • ClassifyAs typical CIDP or a CIDP variant (§3) — the variant changes what treatment works.
  • SupportNCS demyelination (conduction slowing/block, prolonged distal & F-wave latencies, temporal dispersion); CSF (raised protein, normal cells); MRI root/plexus hypertrophy or enhancement; objective response to treatment.

The overdiagnosis trap

  • RealityOnly ~2/3 of patients labelled "CIDP" meet the 2021 criteria (83% sensitivity, 94% specificity) — the criteria were tightened deliberately.
  • MandatoryScreen everyone for a paraprotein (serum electrophoresis + immunofixation + free light chains) — otherwise POEMS, myeloma and anti-MAG neuropathy get missed (§4).
  • RuleIf treatment doesn't work, the diagnosis is the first thing to question — not the dose.
3 Variants — because the treatment response differs
VariantFeaturesTreatment note
Typical CIDPSymmetric proximal + distal sensorimotorResponds to steroids, IVIG or PLEX
Distal (DADS)Distal, sensory-predominant; often IgM paraprotein / anti-MAGLower IVIG response — check the paraprotein
Multifocal (Lewis-Sumner)Asymmetric, multifocal sensorimotor, conduction blockDon't confuse with MMN (pure motor, §4)
FocalConfined to one limb/regionAs typical
MotorPure motorIVIG first-line — steroids may worsen it
SensoryPure sensoryPoor steroid response
4 The look-alikes — mimics & the nodopathies split out in 2021

Mimics to exclude

  • MMNMultifocal motor neuropathy — pure motor, asymmetric, conduction block, anti-GM1. IVIG-responsive but steroids and PLEX make it worse — the mistake to avoid.
  • ParaproteinAnti-MAG IgM (distal, ataxic, poor IVIG response, sometimes rituximab); other MGUS.
  • POEMSPolyneuropathy + organomegaly/endocrinopathy/M-protein/skin; VEGF raised, osteosclerotic myeloma → treat the plasma-cell disorder, not "CIDP".
  • HereditaryCMT — pes cavus, family history, uniform slowing, normal CSF protein.

Autoimmune nodopathies 2021

  • What~10% of "CIDP" have antibodies against nodal/paranodal proteins — CNTN1, NF155, Caspr1, NF140/186. Now a separate entity.
  • Why splitIgG4 subclass, complement-independent — different mechanism, different treatment.
  • PhenotypeAcute/subacute onset, distal weakness, tremor, ataxia, cranial nerve involvement, very high CSF protein, sometimes nephrotic (CNTN1).
  • TreatPoor response to IVIG → rituximab. Test for these antibodies in atypical or IVIG-refractory disease.
5 Treatment

First-line (EAN/PNS)

  • ThreeCorticosteroids, IVIG, or plasma exchange — all effective; no preference between steroids and IVIG; PLEX if the first two fail.
  • SteroidsOffer a real chance of drug-free remission PREDICT — but watch toxicity. Not for motor or sensory variants.
  • Motor CIDPIVIG first-line — steroids can worsen it.

Maintenance & new agents

  • MaintainIVIG, subcutaneous Ig (home-based, steadier levels) PATH, or steroids; minimise the dose and periodically attempt withdrawal — some remit.
  • EfgartigimodSC FcRn inhibitor — reduced relapse ADHERE; first new drug class approved for CIDP (2024).
  • RituximabFor autoimmune nodopathies and refractory disease.

If it isn't working

  • StopNo objective response → question the diagnosis, don't just escalate. Is it a nodopathy, MMN, POEMS, or CMT?
  • MeasureJudge response objectively (grip, INCAT/disability scales), not on symptoms alone — and don't commit someone to lifelong Ig they don't need.
Guidelines & reviews. EAN/PNS guideline on diagnosis and treatment of CIDP — second revision (Van den Bergh et al, Eur J Neurol 2021: typical CIDP vs variants; steroids/IVIG first-line, PLEX if they fail; IVIG/SCIG/steroids for maintenance; autoimmune nodopathies split out); Köller et al, "Chronic Inflammatory Demyelinating Polyneuropathy," NEJM 2005; eTG; PBS.   Trials. ICE (IVIG in CIDP, Lancet Neurol 2008); PATH (subcutaneous immunoglobulin maintenance, Lancet Neurol 2018); PREDICT (pulsed dexamethasone/short prednisolone → remission); ADHERE (subcutaneous efgartigimod, Lancet Neurol 2024 — reduced relapse; basis for 2024 FDA approval, first non-immunoglobulin therapy for CIDP); complement inhibitors (empasiprubart, imeroprubart) in trials.   Caveats. Unlike GBS, steroids work in CIDP — but not in the pure motor or pure sensory variants, and steroids/PLEX worsen MMN. CIDP is overdiagnosed: screen for a paraprotein, and reconsider the diagnosis whenever treatment fails. Autoimmune nodopathies (CNTN1/NF155/Caspr1) are IgG4-mediated, IVIG-refractory, and rituximab-responsive. Confirm PBS status for IVIG/SCIG and efgartigimod. Verify all doses. Companion to the GBS sheet and the neuromuscular set (myasthenia gravis, myopathy, MND).