Updated diagnostic criteria for multiple sclerosis carry a real risk of overdiagnosis and unequal implementation, it is being argued.
The criticisms target the 2024 McDonald criteria, published in The Lancet Neurology [link here], which some clinicians said could result in asymptomatic patients receiving a formal MS diagnosis despite never developing clinical symptoms. However, criteria’s developers have pushed back, defended their approach as flexible and insisting that “a diagnosis does not mandate treatment initiation.”
The RIS problem
A central concern involved patients with radiologically isolated syndrome (RIS), incidental MRI findings consistent with MS in people with no symptoms.
In a letter to the same journal [link here], Professor Darin Okuda of the University of Texas Southwestern Medical Center observed that the new McDonald criteria placed greater emphasis on spinal cord lesions than the separate 2023 RIS criteria. This mismatch, he said, could cut both ways.
Some patients might meet RIS criteria but not receive an MS diagnosis, potentially missing early treatment opportunities. Conversely, asymptomatic patients with sufficient MRI lesions and positive cerebrospinal fluid (CSF) findings could be diagnosed with MS, and some may never develop symptoms, yet face the social stigma, unnecessary treatment and financial consequences of a formal diagnosis.
Professor Stephen Krieger of the Icahn School of Medicine at Mount Sinai raised similar concerns [link here], warning that the updated criteria increased the risk of misdiagnosing asymptomatic patients with non-specific MRI findings. He also flagged the practical challenge of explaining to patients the difference between an MS diagnosis and an elevated risk of developing MS.
Responding in the same issue [link here], the team behind the criteria led by senior author Professor Alan Thompson, honorary consultant neurologist at the UCL Queen Square Institute of Neurology, said the new criteria could identify high-risk asymptomatic patients by incorporating prognostic factors such as sex, age, lesion number and topography, and oligoclonal bands. Treatment decisions, they added, remained part of a patient-doctor process shaped by local regulations and evidence.
Optic nerve findings: expertise required
Professor Ahmed Toosy, also of UCL Queen Square, strongly supported the criteria’s inclusion of the optic nerve as a region for fulfilling dissemination in space (DIS), calling it formal recognition of a region already associated with clinically definite MS. But he warned that correctly interpreting optical coherence tomography, visual evoked potentials or optic nerve MRI required specialist expertise that was not universally available.
Professor Toosy’s group proposed an international Delphi process to develop consensus-driven implementation strategies for optic nerve assessment.
Professor Krieger echoed the concern, warning that variable access to diagnostic tools could drive inconsistent uptake of the criteria and push patients toward specialised MS centres, a trend he said would disproportionately affect patients in low-resource settings and worsen diagnostic delays.
Professor Thompson and colleagues acknowledged that harmonising methodologies and training would be essential, but stressed that advanced techniques were “not mandatory” and could “substantially increase diagnostic confidence when available.”
CSF inflammation should be central, say German authors
Professor Uwe Zettl of Rostock University Medical Center and colleagues argued the updated criteria carried “a substantial risk of confirmation bias,” since MS could now be diagnosed on paraclinical findings alone [link here].
They argued that detection of chronic non-infectious CSF inflammation was “almost pathognomonic” for MS and should serve as the cornerstone of diagnosis once MS was suspected. Other modalities, they said, should confirm or refute that framework rather than independently establish it.
Future revisions, they proposed, should prioritise diagnostic safety over speed, address global applicability, and establish CSF-based confirmation of CNS inflammation as a central diagnostic pillar.
Professor Thompson and colleagues said they had guarded against overdiagnosis by reinforcing the requirement for objective evidence, particularly MRI abnormalities, and incorporating additional safeguards for older patients or those with vascular risk factors, including evaluation of anti-MOG antibodies, the central vein sign and paramagnetic rim lesions.
They welcomed the debate, saying constructive discussion and ongoing education were essential to improve care for people with MS.