Diagnosis

1. Diagnostic Strategy

MS is a clinical diagnosis supported by paraclinical tests — not a diagnosis made by MRI alone. The approach has three pillars:

1. Demonstrate dissemination in space (DIS) — lesions in ≥ 2 of 4 typical CNS regions.
2. Demonstrate dissemination in time (DIT) — lesions of different ages or a second clinical attack; or, under McDonald 2017, the presence of CSF-specific oligoclonal bands.
3. Exclude better explanations — NMOSD, MOGAD, vascular, infectious, sarcoid, vitamin-B12, copper, autoimmune-encephalitis mimics.

2. The McDonald 2017 Criteria

The McDonald criteria, first published in 2001 and revised in 2005, 2010, and 2017 (Thompson et al., Lancet Neurol 2018), are the global standard. Their central innovation is the use of a single MRI to satisfy both DIS and DIT, allowing diagnosis at the first clinical event:

Dissemination in space — ≥ 1 T2 lesion in ≥ 2 of: periventricular, juxtacortical/cortical, infratentorial, spinal cord. (2017 added «cortical» and allowed symptomatic lesions.)

Dissemination in time — (a) simultaneous gadolinium-enhancing and non-enhancing lesions on a single scan, OR (b) a new T2 / gadolinium-enhancing lesion on follow-up MRI, OR (c) CSF-specific oligoclonal bands (a 2017 addition that lets a single scan plus LP make the diagnosis).

3. MRI — T2, Gadolinium, Black Holes, Atrophy

MRI is the dominant paraclinical test in MS. Standard sequences:

• T2/FLAIR — identifies all chronic and acute demyelinating plaques (high signal). T2 lesion load is the principal radiological burden measure.
• T1 post-gadolinium — identifies acute, BBB-disrupting lesions. Enhancement persists ~2–6 weeks. Used to satisfy DIT (simultaneous enhancing + non-enhancing).
• T1 hypointense («black holes») — chronic, axonal-loss lesions; strongest correlate of disability among lesion features.
• 3D-FLAIR + DIR (double inversion recovery) — sensitive for cortical and juxtacortical lesions.
• Susceptibility-weighted imaging — the central vein sign and paramagnetic rim lesions; emerging diagnostic biomarkers.
• Spinal cord — sagittal STIR/PD, axial T2; cord lesions are a strong DIS contributor and a key prognostic factor.
• Brain volumetry — whole-brain atrophy ~0.5–1.0%/yr untreated («PBVC» — percentage brain volume change); thalamic atrophy is the most sensitive regional marker.

4. MAGNIMS Guidelines — the European Imaging Standard

The MAGNIMS network (Wattjes et al., Lancet Neurol 2021) standardised MRI acquisition and reporting in MS. Key elements:

• 3T preferred; minimum sequences 3D FLAIR, 3D T1, sagittal+axial T2 of cord; gadolinium at first scan and when activity-monitoring needed.
• Standardised slice thickness ≤ 3 mm; consistent positioning to allow longitudinal comparison.
• Reporting — lesion count by topography, gadolinium-enhancing lesions, new/enlarging T2 lesions vs prior, T1 hypointensities, atrophy.
• Surveillance frequency — baseline + 6 months for re-baseline post-DMT initiation, then annually if stable; more frequently in early/active disease.

MAGNIMS guidelines now also incorporate the central vein sign as a research-grade specificity tool and PRLs as prognostic markers.

5. CSF — Oligoclonal Bands and the IgG Index

CSF-specific oligoclonal IgG bands (OCBs), first described by Kabat in 1948, are the longest-standing biomarker of MS. The diagnostic test is isoelectric focusing on agarose gel followed by immunoblot with paired serum and CSF samples:

• Type 1 pattern — no bands in CSF or serum (normal).
• Type 2 — CSF bands not present in serum (intrathecal synthesis) — positive for MS; ~95% of MS patients.
• Type 3 — CSF bands plus identical serum bands plus extra CSF-only bands.
• Type 4 — identical CSF and serum bands (passive transfer; systemic disease).
• Type 5 — monoclonal serum component.

Quantitative supporting tests:

CSF white-cell count is typically normal or mildly elevated (< 50 cells/µL); higher counts should prompt re-evaluation for NMOSD, infection, sarcoid, lymphoma, or autoimmune encephalitis.

6. Kappa Free Light Chains (KFLC)

The CSF kappa free light chain (KFLC) index has emerged as an alternative or complement to OCBs:

• Index — (CSF KFLC / serum KFLC) / (CSF albumin / serum albumin); cut-off ~5–10 (assay-dependent).
• Sensitivity / specificity — ~95% / ~90% for MS, comparable to OCBs.
• Quantitative, automated — replaces operator-dependent gel reading; easier to standardise across labs.
• Likely McDonald 2024 acceptance — KFLC index expected to be recognised as equivalent to OCBs in next revision.

7. Evoked Potentials

Evoked potentials measure conduction along central afferent pathways and were major paraclinical tools before MRI. They retain a role in selected settings:

• Visual evoked potentials (VEP) — pattern-reversal stimulus; P100 latency > 115 ms is abnormal; sensitive marker of subclinical optic-nerve demyelination.
• Somatosensory evoked potentials (SSEP) — tibial or median; assess dorsal columns; useful when cord lesions are suspected but MRI ambiguous.
• Motor evoked potentials (MEP) — transcranial magnetic stimulation; central motor conduction time. Increasingly used as a quantitative biomarker.
• Brainstem auditory evoked responses (BAER) — less sensitive than VEP/SSEP in MS; mostly historical.

8. Optical Coherence Tomography (OCT)

OCT non-invasively quantifies the retinal nerve fibre layer (RNFL) and ganglion cell-inner plexiform layer (GCIPL); both are thinned in MS, particularly after optic neuritis but also in non-ON eyes.

• Post-ON RNFL loss — ~10–40 µm reduction over 3–6 months.
• GCIPL — correlates with whole-brain volume; reproducible biomarker.
• Use cases — objective evidence of past ON; monitoring of progression; trial endpoint for remyelinating drugs (clemastine, opicinumab).

9. Misdiagnosis & Red Flags

Misdiagnosis remains common: in tertiary referral series, ~20% of patients carrying an MS label do not have MS (Solomon et al., Neurology 2016). Red flags that should prompt re-evaluation:

• Normal MRI brain in someone with progressive symptoms.
• Longitudinally extensive (≥ 3 segment) cord lesion — think NMOSD, sarcoid, infection.
• Severe bilateral simultaneous optic neuritis — think NMOSD, MOGAD.
• Rapid course (months not years) — consider mimics: vasculitis, B12, sarcoid, lymphoma, ADEM.
• Symmetric, peripheral, length-dependent symptoms — consider CIDP, B12, copper, paraneoplastic.
• Atypical MRI features (large white-matter lesions in basal ganglia, anterior temporal lobes) — consider CADASIL.
• Hyperintense corticospinal tracts — consider hereditary spastic paraplegia, ALS.
• Persistent gadolinium enhancement > 3 months — consider neurosarcoidosis, CNS lymphoma, infection.