CIDP Treatment: Applying EAN/PNS Criteria Before Long-Term IVIG

Why CIDP Treatment Matters

Chronic inflammatory demyelinating polyradiculoneuropathy is one of the few chronic neuropathies that responds to immunotherapy. That makes an accurate diagnosis high stakes in both directions.

A patient with genuine CIDP who is missed accumulates avoidable disability. A patient without CIDP who is labelled with it may face years of infusions, adverse events, and cost for a disease they do not have.

The difficulty is that CIDP has no confirmatory biomarker. The diagnosis is built from a clinical phenotype, electrodiagnostic evidence of demyelination, and the exclusion of mimics — a combination that is easy to misapply.

Because immunoglobulin makes many people feel better regardless of their underlying condition, a positive-seeming response can quietly substitute for a rigorous diagnosis. Applying the current criteria before committing to long-term therapy is the safeguard against that trap.

What the 2021 EAN/PNS Guideline Recommends

The 2021 European Academy of Neurology and Peripheral Nerve Society guideline is the second revision of the earlier EFNS/PNS consensus. It restructured how the diagnosis is framed and how treatment is sequenced.

Two changes matter most for the treatment decision. The guideline replaced the older “atypical CIDP” label with defined CIDP variants (distal, multifocal, focal, motor, and sensory). It also collapsed the previous three certainty tiers into two — CIDP and possible CIDP — because the accuracy of the probable and definite categories did not meaningfully differ.

On therapy, the task force strongly recommended intravenous immunoglobulin or corticosteroids as initial treatment for typical CIDP and its variants, with immunoglobulin (intravenous or subcutaneous) available for maintenance and plasma exchange reserved for inadequate response.

That distinction is the analytic core of this topic. The criteria are meant to be satisfied up front, on phenotype and electrodiagnostic demyelination, with mimics excluded. Retrofitting a diagnosis from a perceived infusion benefit inverts the logic the guideline was written to protect.

Evidence Reviewed

The evidence base for this decision spans three questions: how well immunoglobulin works, how often the CIDP label is wrong or unnecessary, and how the 2021 criteria perform. The table below draws the relevant trials and cohorts together.

All rows verified against a primary source retrieved for this review; every cited figure carries an effect size or count with its denominator. RCT results use causal language; retrospective cohorts are reported as associations and proportions.

Key Findings on CIDP Treatment

Immunoglobulin efficacy is genuine and well replicated

For patients who truly have CIDP, the short-term benefit of immunoglobulin is among the better-established findings in neuromuscular medicine. The ICE trial — the pivotal randomised efficacy trial and the basis for regulatory approval — showed that IGIV-C led to maintained disability improvement in roughly twice as many patients as placebo.

The 2024 Cochrane review, pooling nine randomised trials, reached the same conclusion at the meta-analytic level and found immunoglobulin comparable to intravenous methylprednisolone and oral prednisolone over the same short window. The PATH trial then showed that subcutaneous immunoglobulin can hold that benefit as maintenance therapy.

None of this is in dispute. The efficacy evidence is precisely why the label is so consequential: an effective, expensive, indefinite therapy is being offered on the strength of a diagnosis that is often not re-examined.

How often the diagnosis is wrong

Independent referral cohorts converge on an uncomfortable figure. When patients carrying a CIDP diagnosis were reassessed against criteria, 47% in one series and 45% in another did not have the disease. Broader review estimates suggest around half of patients labelled with CIDP in the United States may not have it.

The recurring drivers are consistent: over-reliance on a subjective sense of benefit from treatment, liberal interpretation of nerve conduction studies as demyelinating, and over-weighting of mild cerebrospinal-fluid protein elevation. In the electrodiagnostic re-analysis, the original interpretation was confirmed in fewer than half of the misdiagnosed studies.

What the criteria buy, and what they cost

The 2021 criteria were tuned toward specificity, which is the property that protects against overdiagnosis. In the Doneddu cohort, specificity for CIDP reached 98%, well above the 2010 criteria. Rajabally and colleagues reported comparably high specificity in a separate cohort.

That gain comes with lower sensitivity — around 74–83% across validation studies — driven largely by stricter electrodiagnostic thresholds and the sensory criteria. The practical reading is not that the criteria are too strict, but that a criteria-negative yet clinically compelling patient warrants supportive testing rather than either reflexive rejection or reflexive lifelong immunoglobulin.

Who still needs long-term therapy

Even among correctly diagnosed patients, indefinite treatment is frequently assumed rather than confirmed. The Adrichem withdrawal trial tested this directly: stable patients were tapered off IVIG under blinded conditions.

Withdrawal was not shown to be non-inferior to continuation, so maintenance remains the safer default for many. Yet 41% of withdrawn patients stayed stable at 24 weeks, and 28% at 76 weeks — consistent with earlier placebo-arm signals that a substantial share of patients are overtreated. A planned reassessment of ongoing need, rather than open-ended continuation, is the evidence-aligned posture.

Quality and Consistency of the Evidence

The strength of the evidence varies by claim, and the certainty language below reflects that.

• High certainty — Short-term efficacy of immunoglobulin in confirmed CIDP. Multiple randomised trials and a current systematic review agree, with a consistent direction and a defined effect on disability.
• Moderate certainty — Frequency of misdiagnosis and overtreatment. The signal is consistent across independent cohorts and one withdrawal RCT, but the underlying data are largely retrospective and single-centre, so exact proportions vary by setting.
• Moderate certainty — The specificity-for-sensitivity trade-off of the 2021 criteria. Several diagnostic-accuracy studies point the same way, though estimates differ with the reference population and the electrodiagnostic protocol used.
• Low certainty — Optimal maintenance dose, interval, and the right moment to attempt withdrawal. No head-to-head trial defines these; practice is extrapolated from trial regimens and expert consensus.

What the Evidence Does Not Show

There is no validated biomarker that confirms CIDP. Diagnosis remains a clinical, electrodiagnostic, and exclusionary judgement, which is why criteria discipline carries so much weight.

The evidence does not settle how best to recover the sensitivity that the 2021 criteria trade away. Nerve ultrasound, MRI, and autoantibody testing are promising adjuncts for criteria-negative but suspicious patients, yet their optimal place in a diagnostic pathway is not defined.

No trial establishes when it is safe to stop maintenance therapy. The withdrawal data identify that overtreatment is common without providing a validated rule for selecting who can taper. Nor is there a standardised objective threshold — which measure, what magnitude of change — for judging ongoing benefit.

Finally, long-term comparative data are limited for immunoglobulin against corticosteroids and against newer mechanisms such as neonatal Fc receptor, complement, and B-cell–directed agents, so maintenance choices rest on short-term and indirect comparisons.

Practical Implications

The evidence supports a diagnosis-first posture before any commitment to long-term immunoglobulin. Five considerations follow from it.

1. Confirm the diagnosis against the 2021 EAN/PNS criteria before starting maintenance therapy. Anchor on clinical phenotype and electrodiagnostic demyelination with mimics excluded — not on an anticipated response to immunoglobulin.
2. Do not let a subjective response stand in for a diagnosis. Judge benefit with objective measures such as the INCAT disability score, grip strength, or a walking assessment, since perceived wellbeing improves in many patients regardless of the underlying condition.
3. Re-read the nerve conduction studies rigorously. Electrodiagnostic misinterpretation is a leading contributor to misdiagnosis; where suspicion is high but the electrodiagnostic criteria are unmet, nerve ultrasound or MRI may add value.
4. Weigh the high specificity but lower sensitivity of the 2021 criteria. A criteria-negative yet clinically compelling case calls for supportive testing and reassessment, not automatic rejection and not automatic indefinite therapy.
5. Build in a planned review of ongoing need. Recheck objective response and, in stable patients, consider a structured taper or withdrawal trial, because a substantial minority remain stable off treatment.

Evidence Grade and Bottom Line

The bottom line is a sequence, not a slogan. Immunoglobulin works for people who have CIDP. A large fraction of people labelled with CIDP do not have it. The 2021 criteria are built to catch that error by favouring specificity. A favourable response to immunoglobulin is a supportive clue, never a diagnosis. And many stable patients are treated longer than they need to be.

Applying the criteria before — not after — committing to long-term IVIG is what keeps effective therapy pointed at the patients who will actually benefit.

Article Information and References