New criteria for multiple sclerosis allow for earlier diagnosis – but experts warn of the risk of overdiagnosis and call for new biomarkers

An international group of leading multiple sclerosis experts believes that the revised 2024 McDonald criteria open up the possibility of diagnosing the disease much earlier and in a wider range of patients, but also create new challenges. In a commentary published on July 9, 2026, in Nature Medicine, 29 experts discuss what needs to happen next: better distinguishing multiple sclerosis from its mimics, learning to predict individual disease progression, and restructuring patient recruitment principles for clinical trials.

A particularly important change is that modern diagnostics rely less and less on waiting for a second clinical episode. The criteria have incorporated new magnetic resonance imaging and laboratory markers, the optic nerve has become the fifth typical anatomical site of involvement, and in some situations, a diagnosis of multiple sclerosis is now possible even in people who have not yet experienced a classic neurological attack. This fundamentally shifts the disease's boundaries to an earlier, sometimes almost preclinical, phase.

Initial data from real-world clinical practice demonstrate how significant the effect can be. In a study of patients referred for evaluation with suspected multiple sclerosis, application of the 2024 criteria increased the number of diagnoses after the initial evaluation from 172 to 220, and the median time to diagnosis decreased from 84 to 40 days. The sensitivity of the new criteria was 92.6%, but the specificity was only 57.8% when using the diagnosis according to the 2017 criteria at the end of follow-up as the reference.

This balance between earlier detection and the dangers of overdiagnosis is the central question of the Nature Medicine article. The authors support the direction of the new diagnostic approach, but emphasize that the ability to diagnose earlier does not necessarily mean the ability to accurately predict who is truly at risk of rapid disease progression, who requires immediate, highly effective therapy, and who may remain minimally active for decades.

What has changed since the McDonald Criteria 2024	Practical significance
The optic nerve has become the fifth typical area of damage.	More ways to prove the spread of the process in the central nervous system
New magnetic resonance biomarkers have been introduced	Specificity of questionable lesions can be increased
Free light chains of kappa in cerebrospinal fluid were added.	Additional laboratory confirmation
Some patients without a classic attack may be diagnosed with	The diagnostic boundary shifts to the preclinical phase
When a large number of typical areas are affected, the requirements are simplified	The diagnosis can be made faster
Special recommendations have been issued for the elderly and patients with underlying medical conditions.	An attempt to reduce misdiagnosis
The main unresolved problem	Prognosis of the individual course of the disease

Why the Nature Medicine article is not a new clinical trial

The publication is a Commentary, meaning it is an expert analysis, not a new randomized trial or a new patient cohort. Its authors are specialists in neurology, neuroimaging, neuroimmunology, and clinical research from leading centers in Europe, North America, Asia, and Australia. Among them are Jiwon Oh, Frederik Barkhof, Amit Bar-Or, Massimo Filippi, Daniel Reich, Xavier Montalban, Mar Tintore, and Peter Calabresi.

The article appeared after the official publication of the 2024 McDonald criteria. Despite the title "2024," the final peer-reviewed version of the revised criteria was published in The Lancet Neurology in September 2025 and appeared in the October issue of the journal. Therefore, the new Nature Medicine commentary effectively represents the next step: the authors discuss not which rules to adopt, but what scientific and clinical implications have arisen since their adoption.

The criteria have changed significantly because the understanding of multiple sclerosis has become significantly more complex in recent years. The disease can no longer be conveniently described solely as a series of discrete inflammatory attacks followed by a later "progressive phase." Current data indicate that inflammatory activity, latent nervous system damage, and progression can coexist virtually from the onset of the disease.

Therefore, the Nature Medicine article addresses the issue beyond diagnostics. If medicine has learned to detect a disease before significant disability occurs, it is also necessary to learn to determine its biological activity, the likelihood of its progression, and the expected response to treatment. Otherwise, more sensitive diagnostics could lead to a situation where diagnoses become more accurate and earlier, but doctors still have difficulty explaining to a specific patient what this diagnosis means for the next 10-20 years.

Publication characteristics	Data
Magazine	Nature Medicine
Publication type	Comment
Date of publication	July 9, 2026
Main theme	Consequences of the 2024 McDonald Criteria
New clinical trial	No
The main task	To identify unsolved problems in diagnosis, prognosis and research
Authors	International Expert Group on Multiple Sclerosis
DOI	10.1038/s41591-026-04490-8

What were the old McDonald criteria?

The diagnosis of multiple sclerosis has historically been based on a simple concept, yet challenging in practice: demonstrating that the inflammatory-demyelinating process affects different areas of the central nervous system and occurs at different points in time. These two principles are known as "dissemination in space" and "dissemination in time."

In the era before modern magnetic resonance imaging, this often meant waiting for a second neurological attack. For example, a person might initially develop optic neuritis, followed months or years later by sensory disturbances or weakness consistent with the new lesion. Only the accumulation of such events allowed one to confidently diagnose a widespread central nervous system disorder.

Magnetic resonance imaging has gradually replaced the expectation of clinical events with visual evidence. The 2017 criteria already allowed for the use of the characteristic distribution of lesions and the appearance of new lesions in subsequent imaging studies. Furthermore, specific oligoclonal immunoglobulin bands in cerebrospinal fluid could, in certain circumstances, replace evidence of dissemination over time.

But even after the 2017 revision, there were still patients whose biological process likely already represented multiple sclerosis, although the formal diagnostic threshold had not yet been reached. It is precisely this "gray area" that the new criteria attempt to narrow by using more specific features from magnetic resonance imaging, cerebrospinal fluid, and visual system examinations.

Concept	What does it mean?
Dissemination in space	The lesions are present in several typical areas of the central nervous system.
Dissemination in time	There is evidence of the occurrence of inflammatory lesions at different periods
Clinical attack	A new typical neurological episode associated with demyelination
Clinically isolated syndrome	First typical clinical episode without full fulfillment of previous criteria
Radiologically isolated syndrome	Typical lesions of multiple sclerosis were found without classic symptoms
The main trend of the criteria	Reducing the need to wait for the next clinical event

The optic nerve is now officially the fifth site of damage.

One of the most notable changes is the addition of the optic nerve as a fifth anatomical region of the central nervous system, which can be used to demonstrate spatial dissemination of the process. Previously, the four main regions were considered to be the periventricular, cortical or juxtacortical, infratentorial, and spinal cord regions.

The change seems logical: optic neuritis is one of the classic manifestations of multiple sclerosis. However, previously, optic nerve damage did not have the same formal status in the criteria as a characteristic lesion in the brain or spinal cord. Now, objectively confirmed optic nerve damage can help meet the spatial dissemination criterion.

Confirmation of damage can be achieved not only by symptoms. Retinal optical coherence tomography and visual evoked potentials are additional tools. Recommendations include, for example, interocular differences in the thickness of the peripapillary retinal nerve fiber layer of at least 6 micrometers, or differences in the thickness of the ganglion cell layer and inner plexiform layer of at least 4 micrometers, provided there is no other explanation for the asymmetry and strict quality control of the examination has been performed.

Visual evoked potentials (VEPs) can detect slowing of electrical signal conduction along the optic nerve, characteristic of demyelination. However, the authors of specialized guidelines emphasize that the results depend on the equipment, methodology, and laboratory, and therefore cannot be interpreted mechanically using a single, universal number.

Inclusion of the optic nerve is especially important for patients whose first manifestation of the disease was optic neuritis. Previously, the physician sometimes had to search for a sufficient number of additional lesions or wait for a new event; now, objective damage to the optic nerve itself becomes part of the anatomical map of the disease.

Typical area	It was taken into account earlier	2024 Criteria
Periventricular	Yes	Yes
Cortical/juxtacortical	Yes	Yes
Infratentorial	Yes	Yes
Spinal cord	Yes	Yes
Optic nerve	No, as a separate topography	Yes

Some patients can now receive a diagnosis without waiting for the disease to spread over time.

An even more fundamental change concerns the classic requirement for temporal dissemination. In some clinical situations, the volume of compelling spatial and biomarker data is now considered so vast that waiting for a new lesion or a second clinical attack is no longer necessary. This marks a shift from a strictly chronological definition of the disease to a more biological approach.

In particular, if typical lesions are present in four or five characteristic anatomical regions in a patient with a suitable clinical presentation, separate confirmation of temporal dissemination may no longer be necessary. The less pronounced the spatial spread, the more important additional laboratory or magnetic resonance imaging evidence becomes.

From a clinical perspective, the benefit is clear. In highly active multiple sclerosis, waiting several months for a second event can mean additional irreversible damage to nerve tissue. If the available data already almost certainly points to the disease, the need to artificially wait for its next manifestation becomes difficult to justify. The authors of Nature Medicine cite earlier and more inclusive diagnosis as one of the main achievements of the revision.

But at the same time, the cost of diagnostic error is changing. The earlier and more widely the definition is applied, the more people are on the borderline between true early-onset multiple sclerosis and diseases that can mimic its magnetic resonance imaging. Therefore, the new criteria simultaneously require more specific biomarkers—a central vein within the lesion, a paramagnetic rim, and immunological signatures in the cerebrospinal fluid.

Old logic	New logic
It is necessary to prove the defeats in space and time	In certain situations, a large amount of specific data can replace waiting
The second attack had great diagnostic significance.	Diagnosis is increasingly possible after the first event.
MRI mainly showed the number and location of lesions	The internal structure of some foci is also taken into account.
Cerebrospinal fluid used mainly oligoclonal bands	Added free light chains kappa
Risk of late diagnosis	It's decreasing
Potential risk of overdiagnosis	May increase

The first real comparison showed that diagnosis was made approximately twice as fast.

Highly significant data was published in Neurology in 2026. Researchers analyzed 347 patients referred to a specialized center in 2024 with clinical or radiological suspicion of multiple sclerosis. The average age was 39.4 years, 66% of participants were women, and the average follow-up lasted approximately 15.3 months.

In 73 patients, or 21%, alternative diagnoses were ultimately identified and excluded from the main diagnostic analysis. Among the remaining 274 patients after the initial assessment, the 2017 criteria allowed a diagnosis of multiple sclerosis to be made in 172 patients, while the 2024 criteria allowed a diagnosis of multiple sclerosis to be made in 220 patients. The difference was statistically significant.

By the end of the follow-up, the difference remained: 204 people were diagnosed according to the 2017 version and 237 according to the 2024 version. At the same time, the number of people remaining in the clinically or radiologically isolated syndrome categories decreased. The median time from the first clinical examination to diagnosis decreased from 84 to 40 days.

The sensitivity of the 2024 criteria relative to the diagnosis established according to the 2017 criteria by the end of the follow-up period was 92.6%, and the overall accuracy was 83.6%. This means that the new rules significantly less frequently missed patients who later met the traditional definition of the disease.

However, the specificity was only 57.8%. The study's authors emphasize that the results should be interpreted with caution: the old criteria themselves are not an absolute biological "gold standard," and the follow-up was relatively short. Furthermore, the study lacked complete information on the central venous signature and kappa free light chains, which could potentially improve the specificity of the new system.

Indicator	Criteria 2017	2024 Criteria
Diagnosis after initial examination	172	220
Diagnosis by the end of observation	204	237
Median time to diagnosis	84 days	40 days
Sensitivity of new criteria	-	92.6%
Accuracy	-	83.6%
Specificity	-	57.8%
The main advantage	-	Earlier detection
The main question	-	How to maintain specificity

Multiple sclerosis can now sometimes be diagnosed even before classic symptoms appear.

One of the most controversial conceptual shifts has been the conflation of multiple sclerosis and radiologically isolated syndrome. This refers to a situation where magnetic resonance imaging incidentally reveals lesions very similar to demyelination seen in multiple sclerosis, even though the individual has never had a characteristic clinical attack.

Previously, such a person was not formally considered a patient with multiple sclerosis based solely on the characteristic magnetic resonance imaging (MRI) pattern. The 2024 criteria allow for the diagnosis in some such cases if the spatial distribution of lesions is combined with additional compelling features—for example, characteristic activity over time, positive CSF markers, or highly specific MRI features.

On the one hand, the biological logic of this decision is clear. A clinical symptom appears only when the damage has reached a certain functional threshold or affected a critical area. The absence of a symptom does not mean the absence of an inflammatory process. Modern longitudinal studies of multiple sclerosis show that a significant portion of magnetic resonance activity can occur without new neurological complaints.

For example, a large analysis of over 8,000 patients showed that over 90% of the activity detected during routine magnetic resonance imaging (MRI) of the brain and spinal cord may not be accompanied by new or worsening clinical symptoms. This subclinical activity was nevertheless associated with subsequent damage accumulation and progression to more severe disease states.

On the other hand, not every person with radiologically isolated syndrome will necessarily develop clinically significant disease within the same timeframe. Therefore, shifting the diagnostic label to the preclinical stage immediately raises a second question: which of these individuals should be treated. This is why the authors of Nature Medicine link the new diagnostics to the urgent need for prognostic biomarkers, not just diagnostic ones.

State	Clinical symptoms	Typical foci
Radiologically isolated syndrome	There is no classic attack	Eat
Clinically isolated syndrome	One typical episode	Often there is
Multiple sclerosis according to the previous criteria	More evidence of distribution was usually required	Eat
2024 Criteria	In some cases, they allow for an earlier diagnosis	Additional specific biomarkers are taken into account

A central vein within a lesion helps distinguish multiple sclerosis from its mimics

A common white spot on MRI is not a unique sign of multiple sclerosis. Similar changes can occur with migraines, vascular diseases, aging, and other inflammatory processes. Therefore, increasing diagnostic sensitivity without increasing specificity can be problematic.

One new tool is the central vein sign. In multiple sclerosis, the inflammatory focus often forms around a small venule, so on specialized magnetic resonance imaging (MRI) sequences, the vessel can be seen running through the center of the lesion. This reflects the characteristic perivenular biology of the disease, rather than simply the shape of a white spot on the image.

The inclusion of the central venous signature changes the diagnostic philosophy of magnetic resonance imaging. Previously, the physician primarily asked, "How many lesions are there and where are they located?" Now, the question of "What is the structure of the lesion itself?" is becoming increasingly important. This makes it possible to potentially differentiate pathological changes that appear similar on standard sequences.

The availability of appropriate sequences and the quality of interpretation remain practical limitations. Specialized magnetic resonance imaging is not performed uniformly everywhere, and the criteria must be valid not only in leading academic centers with high-end equipment, but also in regular hospitals across the globe. This is why the authors of the new article emphasize the need for further standardization of biomarkers.

MRI sign	What does it show?
Normal T2 lesion	An area of altered tissue, but not necessarily specific to MS
Central venous sign	A small vein passing through a characteristic lesion
Potential advantage	Increasing specificity
The main problem	Requirements for MRI protocol and interpretation
Role in the 2024 criteria	Supporting specific evidence

The paramagnetic rim shows chronically active lesions

The second new magnetic resonance marker is paramagnetic rim lesions. Some chronic multiple sclerosis lesions retain activated iron-containing immune cells at the periphery. On magnetic susceptibility-sensitive sequences, this area can appear as a characteristic dark rim.

Such a lesion differs biologically from a completely inactive scar. It reflects ongoing, "smoldering" inflammation at the interface of the old lesion and the surrounding tissue. Therefore, the paramagnetic rim is of interest not only as a way to confirm that the lesion is typical of multiple sclerosis, but also potentially as an indicator of chronic disease activity.

This is especially important given the changing understanding of progressive multiple sclerosis. Previously, progression was often contrasted with acute relapses: the disease was thought to be initially inflammatory and later become predominantly neurodegenerative. Current data show a much greater overlap: localized inflammation and slowly expanding lesions can already exist in the early stages.

However, paramagnetic rim lesions cannot yet be considered a universal predictor of future disability in a specific individual. Their standardized detection depends on equipment and protocols, and the significance of the number and dynamics of such lesions still requires long-term verification. The authors consider this transition from a diagnostic marker to a reliable prognostic marker to be one of the main future challenges.

Free kappa light chains expand cerebrospinal fluid analysis capabilities

In the diagnosis of multiple sclerosis, cerebrospinal fluid analysis has long been used to detect signs of immune activity directly within the central nervous system. The most well-known test is specific oligoclonal bands of immunoglobulins, which are not found in the blood but are detected in the cerebrospinal fluid.

The 2024 criteria added another indicator: free kappa light chains. These are produced by B lymphocytes and plasma cells during antibody synthesis. Increased intrathecal production of these chains can serve as a quantitative indicator of immune activity within the central nervous system.

The practical advantage lies in the possibility of automated quantitative analysis. Oligoclonal bands require specialized laboratory testing and visual pattern interpretation, whereas the kappa free light chain index is potentially easier to standardize and use in a larger number of laboratories. However, thresholds and laboratory methods must be validated.

The new criteria consider free light chain kappa not as a standalone "MS test" capable of independently establishing a diagnosis, but as part of a comprehensive diagnostic system. The results should be interpreted in conjunction with the clinical picture, magnetic resonance imaging, and the exclusion of alternative causes of central nervous system inflammation.

CSF biomarker	What does it reflect?
Oligoclonal bands	Local production of immunoglobulins in the central nervous system
Free light chains of kappa	Intrathecal activity of B-cell immune response
The main advantage of the new marker	Possibility of quantitative analysis
Can one test make a diagnosis?	No
Role	Additional evidence in the general diagnostic system

The main danger of an earlier diagnosis is the mistaken diagnosis of another disease as multiple sclerosis.

Multiple sclerosis does not have a single definitive diagnostic test, such as a specific mutation or the presence of a specific microorganism. The diagnosis remains complex: characteristic symptoms, magnetic resonance imaging, cerebrospinal fluid examination, and the exclusion of diseases that can produce similar symptoms. Therefore, each expansion of the criteria increases the burden of differential diagnosis.

Interpreting white spots in older adults is particularly challenging. Hypertension, diabetes, smoking, dyslipidemia, and other vascular factors can cause multiple white matter changes that visually resemble demyelination. For this reason, the new criteria include specific caveats for individuals aged 50 and older and those with vascular comorbidities.

Early real-world data confirm the need for such caution. In the Neurology study, among 88 patients over 50, 31, or 35%, ultimately had an alternative diagnosis and were excluded. Among the remaining patients, application of the 2024 criteria increased the number of diagnoses after the initial examination from 33 to 46. Moreover, all elderly patients diagnosed with multiple sclerosis had spinal cord pathology, a positive cerebrospinal fluid (CSF), or both.

This well illustrates the central principle: more sensitive criteria should not become an automatic algorithm of "see a few blind spots and diagnose multiple sclerosis." On the contrary, the more atypical the age, symptoms, or underlying conditions, the more additional confirmatory features should be sought.

Situation	Why special care is required
Age ≥50 years	Vascular lesions in white matter are more common.
Arterial hypertension	May lead to microangiopathic changes
Diabetes	Increases vascular load
Migraine	May be accompanied by non-specific white lesions
Smoking and dyslipidemia	Additional vascular factors
Atypical symptoms	Increase the likelihood of another disease
What helps?	Spinal cord, CSF, CVS, PRL and thorough clinical evaluation

A quick diagnosis does not answer the patient’s main question: “What will happen to me next?”

Diagnostic criteria primarily answer a binary question: does the data meet the definition of multiple sclerosis? But for the patient, other questions are far more important: how quickly the disease will progress, whether it will remain asymptomatic, whether disability will occur, and what therapy should be chosen. It is here, according to the authors of Nature Medicine, that modern medicine still falls significantly short of diagnostics.

Two patients may simultaneously meet the same diagnostic criteria but have completely different disease biology. One may constantly develop new lesions and rapidly accumulate structural damage, while another's magnetic resonance imaging may remain virtually unchanged for years. The diagnosis of multiple sclerosis itself does not capture this difference.

Existing clinical markers—the number of relapses, new magnetic resonance lesions, and changes in disability scores—help assess activity, but often respond only after damage has occurred. Biomarkers capable of early detection of the risk of progression and occult damage to the nervous system are needed. The authors of Nature Medicine cite the improvement of such biomarkers as a key requirement for realizing the potential of the new criteria.

Potential candidates include blood measures of neuroaxonal damage, immunological markers, characteristics of chronically active magnetic resonance lesions, quantitative atrophy of the brain and spinal cord, and combined models. However, diagnostic markers and prognostic markers are not the same thing: a marker may be highly confirmatory of multiple sclerosis but say little about the rate of future disability.

The old division of the disease into "recurrent" and "progressive" forms is becoming too crude

Multiple sclerosis is traditionally divided into relapsing-remitting, secondary progressive, and primary progressive variants. These categories have been used for decades to describe the disease, register drugs, and recruit patients for clinical trials.

However, analysis of large data sets shows that the biology of the disease does not fit neatly into three isolated boxes. A Nature Medicine 2025 study analyzed data from 8,023 patients, 118,235 clinical visits, and over 35,000 magnetic resonance imaging scans from nine clinical trials. The results were further validated against an independent sample of 2,243 participants and a real-world cohort of 2,080 patients.

The algorithm identified four key dimensions of the disease: physical disability, cumulative brain damage, clinical relapses, and asymptomatic magnetic resonance activity. Patients with traditionally diverse diagnoses often shared similar biological states, and their clinical course more closely followed a continuum from early/mild disease to advanced stages.

Crucially, asymptomatic MRI activity was not a neutral state. In the model, the probability of transitioning directly from a state of marked asymptomatic activity to advanced disease was approximately 11%, while after a clinical relapse, it was approximately 18%. Each episode of inflammatory activity increased the likelihood of subsequent accumulation of more severe damage.

The model predicted individual progression toward an advanced state with a concordance score of approximately 0.82, which is encouraging but not yet sufficient for automated therapeutic decision-making. The authors of the current Nature Medicine commentary view such biological stratification as a direction that should be developed in parallel with the expansion of diagnostic criteria.

Traditional classification	A newer biological model
Relapsing-remitting MS	Early/mild/evolving state
Secondary progressive MS	May include varying levels of activity and damage
Primary progressive MS	May overlap biologically with other forms
Main classifier	Clinical course of the disease
New model	Disability + brain damage + relapses + latent MRI activity
main idea	MS is a continuum, not three completely separate diseases.

Why this changes clinical drug trials

Most clinical trials of multiple sclerosis have been built around established diagnostic categories. One study recruited people with the relapsing form, another with the primary progressive form, and a third with the secondary progressive form. However, if new criteria change the time of diagnosis and include some patients who previously fell into the clinically or radiologically isolated syndrome categories, the composition of the study populations also begins to change.

A person who would have been diagnosed in a year under the 2017 criteria could be diagnosed today under the 2024 criteria. If such patients are included in clinical trials alongside people with longer-standing and more active disease, the average relapse rate, progression rate, and number of new lesions in the control group may change. This potentially impacts the calculation of the required trial size and the interpretation of treatment effectiveness.

An even more significant problem arises in preclinical multiple sclerosis. If some patients with radiologically isolated syndrome now receive a formal diagnosis, it becomes possible to conduct preventive treatment trials before the first neurological attack. However, such trials require endpoints that reflect real clinical benefit, not just a change in diagnostic label.

Therefore, the authors believe it is necessary to simultaneously modernize diagnostic criteria, the disease stratification system, and the design of clinical trials. Earlier diagnosis creates a unique opportunity to intervene before significant irreversible damage to the nervous system occurs, but success will depend on the ability to correctly select patients for whom the expected benefit of treatment outweighs its risks.

New drugs further blur the line between inflammation and progression

Modern treatment research also shows that the old dichotomy between "inflammatory" and "progressive" disease may be overly simplistic. For example, in the HERCULES study, the Bruton tyrosine kinase inhibitor tolebrutinib was studied in patients with non-relapsing secondary progressive multiple sclerosis (SPMS)—a group for which therapeutic options have traditionally been particularly limited. This study is among the studies cited in the Nature Medicine commentary.

In two GEMINI studies, the same drug was compared with teriflunomide in patients with relapsing multiple sclerosis. Tolebrutinib was not superior to teriflunomide in reducing the annual relapse rate. The different results across different disease components highlight that a single drug can target processes that are poorly captured by traditional disease type classifications.

Such findings stimulate the search for new endpoints, particularly for progressive disease. If disability can worsen regardless of overt relapse, counting clinical attacks alone becomes insufficient. It is necessary to measure underlying inflammatory activity, chronic brain damage, neuroaxonal loss, and functional decline.

This is why the Nature Medicine article links the diagnostic revision to a broader transformation of the entire concept of multiple sclerosis. The boundary between "disease" and "no disease" has become more precise and shifted earlier, but the next challenge is transforming the diagnosis into a multidimensional description of the biological state of a specific individual.

What the new criteria mean for children and elderly patients

One of the goals of the revision was to create a unified approach across the lifespan, rather than rules developed primarily for young adults with typical relapsing disease. The official publication of the criteria explicitly states that they are intended for patients from pediatric age to late onset.

This is important for children, as early inflammatory disease of the central nervous system must be distinguished from post-infectious demyelinating conditions and other childhood neuroimmunological diseases. Moreover, the evidence and typical magnetic resonance imaging may differ from those in the adult population, especially in young children.

In a clinical efficacy study of the 2024 criteria among 27 children with suspected multiple sclerosis, the new rules were sufficient for a slightly higher number of diagnoses, and sensitivity remained high. However, the size of this subgroup was very small, so definitive conclusions about pediatric specificity cannot yet be drawn from these data.

For people over 50, the problem is the opposite: nonspecific vascular changes in the brain are becoming increasingly common. Therefore, with late onset, the criteria urgently require additional confirmatory features and careful exclusion of other causes. An early diagnosis should mean a more accurate diagnosis, not simply a lower threshold for assigning a disease label.

Group	The main diagnostic problem
Children	Distinguish MS from other inflammatory diseases of the central nervous system
Young adults	The most typical MS population
≥50 years	Non-specific vascular MRI lesions are significantly more common
Vascular risk factors	Complicates the interpretation of white matter in the brain
General Principle of the 2024 Criteria	A unified approach with additional precautions

Early diagnosis may allow for earlier protection of the brain's "reserve"

One argument in favor of the earliest possible diagnosis is the concept of the brain's functional and structural reserve. The central nervous system is capable of compensating for damage over a long period of time, so new lesions and loss of nerve fibers can occur without obvious deterioration in well-being. When the reserve diminishes, the same amount of new damage begins to produce significantly more pronounced symptoms.

A large-scale analysis of disease trajectories showed that cumulative magnetic resonance imaging (MRI) load—T2 lesion volume and degree of atrophy—is associated with a lower likelihood of subsequent functional recovery and a higher risk of further progression. Therefore, the authors propose that the accumulation of brain damage itself should be considered an important component of disease worsening.

In this analysis, disease-modifying drugs reduced the likelihood of patients transitioning from an early stage to states of pronounced asymptomatic magnetic resonance activity and clinical relapse. As a result, patients remained in the earlier and less severe portion of the disease spectrum for longer. This supports the biological logic of early suppression of activity, although the specific choice of therapy depends on the individual clinical situation.

However, early treatment is only justified when the diagnosis is sufficiently reliable and the expected risk of the disease truly outweighs the risk of treatment. Some highly effective drugs can increase the risk of infections and other complications, and treatment often lasts for years. Therefore, the main goal of future research is not simply to diagnose everyone earlier, but to accurately identify those who will benefit most from early treatment.

A diagnostic biomarker and a prognostic biomarker are fundamentally different things.

A central venous sign can help us say, "This lesion looks more like multiple sclerosis than a vascular lesion." Free light chains (kappa) help reveal immune activity within the central nervous system. A paramagnetic bezel can identify a specific type of chronically active lesion. But none of these tests alone can yet provide a definitive answer to the question of a specific patient's future.

A prognostic biomarker must address a much more complex challenge: predicting the likelihood of disability, damage accumulation, or progression over years. Moreover, it should ideally be effective even before these dynamics become clinically evident. The lack of such a reliable prognostic tool remains one of the major gaps in modern MS treatment.

An additional challenge is predicting the response to a specific therapy. It's not enough for a patient to know that their disease is potentially active; they need to understand which of the available disease-modifying therapies is most likely to provide the optimal balance of efficacy and safety. Currently, clinical decisions are based on a combination of disease activity, age, comorbidities, reproductive plans, MRI results, and the physician's experience.

Therefore, the authors propose viewing the 2024 criteria as the beginning of a new phase, rather than a final solution to the diagnostic problem. The next goal is to move from the question "Does a person have multiple sclerosis?" to "What specific biological variant of the disease is developing?", "How dangerous is it?", and "What is the optimal treatment right now?"

Marker type	What issue should be resolved?
Diagnostic	Is this really multiple sclerosis?
Activity marker	Is the disease active now?
Prognostic	What is the risk of future deterioration?
Predictive	Which therapy is the patient most likely to respond to?
Damage marker	How much nerve tissue has been lost already?
The main need of the future	Combine this data into a personalized profile

What does the article mean for a patient who is only suspected of having multiple sclerosis?

The first practical consequence is that the diagnostic process may indeed become shorter. If the clinical presentation and magnetic resonance imaging (MRI) findings are characteristic, additional modern biomarkers may allow the physician to make a diagnosis after the first episode, without waiting for new symptoms or a new MRI scan several months later.

Second, a mere set of white spots on an MRI scan still does not constitute a diagnosis. The new criteria do not eliminate the need to exclude alternative diseases. On the contrary, the emergence of more specific features demonstrates the importance of distinguishing between true demyelinating lesions and nonspecific changes.

Third, an earlier diagnosis may occur in a person with minimal symptoms or even without a classic neurological attack. This situation is psychologically and medically more challenging than a traditional diagnosis after several relapses, because the person must decide on long-term observation or treatment before significant limitations develop.

Finally, the diagnosis itself does not automatically predict the severity of the disease's future course. It is important for patients to distinguish between "diagnostic criteria are met" and "the disease will inevitably progress rapidly." This lack of accurate individual prognoses is one of the main reasons the authors call for the development of next-generation biomarkers.

What the new criteria allow - and what they don't allow yet

The 2024 criteria allow	They don't allow it yet
Confirm RS faster	Accurately predict the rate of progression
Use the optic nerve as a fifth area	Consider any visual impairment as a sign of MS
Use CVS and PRL	Completely eliminate all mimic diseases
Use free light chains of kappa	Make a diagnosis based on just one test
In some cases, MS can be diagnosed without a second attack.	To claim that every such patient needs the same treatment
In some cases, diagnose a preclinical disease	Know exactly when a person will develop symptoms
Unify diagnostics for different ages	Resolving difficulties in elderly patients
Create an earlier point for therapy	Determine the ideal drug for each person

The main limitations of the new concept

The first limitation is that most of the criteria are based on a combination of studies with different designs and expert consensus. Their true diagnostic efficacy is only beginning to be tested in broad clinical populations. The first study, scheduled for 2026, shows promising sensitivity, but the moderate specificity suggests that definitive validation requires longer-term follow-up and independent cohorts.

Second, the most innovative elements of the criteria are not yet available everywhere. The central venous signature and paramagnetic rims require suitable magnetic resonance imaging (MRI) sequences and specialists skilled in their interpretation. Kappa free light chain analysis is also not yet standardized across all laboratories.

Third, criteria extend the diagnosis to earlier stages faster than science has learned to determine the prognosis for these stages. This creates the risk of "diagnostic preemption": a doctor can confidently diagnose a disease, but have significantly less confidence in what will happen without treatment and what level of therapeutic intervention is justified.

Fourth, the results of clinical trials conducted under previous disease definitions cannot be automatically generalized to all patients who now meet the expanded 2024 criteria. A person with asymptomatic early disease may differ biologically and in absolute risk from a participant in a traditional trial with multiple clinical relapses. This is one reason why the authors call for adaptations in the design of future trials.

Conflicts of interest and funding

The publication is an international expert commentary, and a significant number of its authors are both involved in multiple sclerosis clinical trials and consult for pharmaceutical companies. Nature Medicine has published a detailed disclosure of their financial relationships with drug manufacturers, including Biogen, Roche, Novartis, Sanofi, Bristol Myers Squibb, Merck, and others.

For example, Jiwon Oh reported consulting or teaching relationships with Amgen, AstraZeneca, Biogen, Bristol Myers Squibb, EMD Serono, Novartis, Roche, and Sanofi Genzyme, as well as research funding from a number of organizations. Similar academic and industrial relationships were declared by many other authors.

This is especially important to consider, as expanding diagnostic criteria potentially increases the number of people who may become candidates for disease-modifying treatments earlier. The mere presence of conflicts of interest does not mean the proposed concept is incorrect, but it makes transparency and independent validation of criteria particularly important.

The study also received support from the National Institutes of Health's Intramural Research Program for participating NIH investigators. The authors emphasize that the findings of this publication do not necessarily reflect the official position of the National Institutes of Health or the U.S. Department of Health and Human Services.

The main conclusion

The 2024 revision of the McDonald criteria marks a shift from a model in which physicians often had to wait for the disease to manifest itself with a new attack to a model capable of detecting multiple sclerosis based on a combination of earlier and more specific biological features. The optic nerve, the structure of magnetic resonance lesions, and new cerebrospinal fluid markers significantly expand diagnostic capabilities.

Initial clinical data indeed show a significant improvement: the median time to diagnosis has decreased from 84 to 40 days, and the number of patients receiving a diagnosis after the initial examination has increased significantly. At the same time, the moderate specificity in the early study serves as a reminder that expanding the disease's boundaries requires careful exclusion of other diagnoses.

But the authors of Nature Medicine believe the main challenge lies in the next stage. Modern medicine has become increasingly adept at detecting the onset of a disease, but remains significantly less adept at predicting its individual course. Diagnosis must gradually evolve from a binary label to a biological profile, including inflammatory activity, the extent of damage already accumulated, the risk of progression, and the expected response to specific therapy.

So the news isn't just that multiple sclerosis can now be diagnosed earlier. A deeper transformation lies in the attempt to redefine the disease itself as a continuous biological process that begins before symptoms manifest and changes over time, and then to restructure diagnosis, treatment, and clinical trials around this new model.

News source

Oh J., Barkhof F., Bar-Or A., Ciccarrelli O., Coetzee T., Filippi M., Granziera C., Hacohen Y., Hemmer B., Kantarci OH, Kim HJ, Lebrun-Frenay C., Montalban X., Mowry E., Ontaneda D., Prat A., Reich DS, Rovira A., Sati P., Siva A., Solomon AJ, Sormani MP, Stankoff B., Thompson A., Tintore M., Traboulsee A., van der Walt A., Wiendl H., Calabresi PA Challenges and future directions for multiple sclerosis after the 2024 McDonald diagnostic criteria. Nature Medicine. Published July 9, 2026.

The publication is categorized as a Commentary, meaning it presents an expert analysis of the implications of the new diagnostic criteria rather than a standalone clinical trial. Its key conclusion is that the 2024 criteria are an important step toward earlier and more inclusive diagnosis, but realizing their potential requires improvements in biomarkers, disease stratification, and clinical trial design.

The main document analyzed by the authors is Montalban X. et al. Diagnosis of multiple sclerosis: 2024 revisions of the McDonald criteria. The Lancet Neurology. 2025;24(10):850–865. Its DOI: 10.1016/S1474-4422(25)00270-4.

DOI of the main article Nature Medicine: 10.1038/s41591-026-04490-8.