Comparison of the Performance of the 2024 and 2017 McDonald Criteria for the Diagnosis of Multiple Sclerosis in Real Life

Overview

The new diagnostic criteria for multiple sclerosis (MS) were published in October 2025 (Montalban et al. 2025). They introduce significant changes compared to the previous criteria from 2017: inclusion of previously "borderline" forms and use of new imaging techniques, inclusion of ophthalmological data, and a new cerebrospinal fluid (CSF) test: the kappa index. It is important to document whether these 2024 criteria, when applied in real life, will enable the diagnosis of patients who would not have been diagnosed using the previous criteria. Furthermore, the threshold to be used for the kappa index in relation to the reference test (IgG oligoclonal bands Obs in CSF) is insufficiently documented. Primary objective: to compare the percentage of patients diagnosed with MS following initial neurological assessment according to the 2024 McDonald criteria and the previous 2017 version. Secondary objectives: compare the percentage of patients receiving the diagnosis according to the two versions; document the need for the new tests appearing in the 2024 criteria; determine the optimal threshold for the kappa index. Methodology: all patients consulting in the neurology department of the Nancy and Metz hospitals between September 1, 2025, and December 31, 2026, followed for two years. No additional tests beyond those required for routine care. The 2017 version of the diagnostic criteria does not require any other tests. The diagnosis according to the 2017 criteria can therefore be reconstructed for each patient. At the end of the initial assessment, the percentage of patients diagnosed according to the two versions is compared, and this comparison will be repeated at different points in time over a two-year period. The contribution of each test will be evaluated using standard diagnostic performance markers, and the relevant kappa index threshold for reaching a diagnosis will be evaluated in relation to the presence of IgG OBs (gold standard). The expected outcomes are the use of the most useful tests to achieve early diagnosis without complicating management, and to estimate which tests are not relevant for diagnosis in order to enrich the knowledge that will enable future revision of these criteria.

Context After initial symptoms suggestive of multiple sclerosis (MS), once neurological and non-neurological differential diagnoses have been ruled out, the question is whether the inflammation in the central nervous system (CNS) will progress and lead to permanent disability. Strictly speaking, the term "multiple sclerosis" can only be used when there is certainty that the inflammation will persist, with the potential to affect the entire CNS in the years to come. By definition, MS is a chronic progressive disease, as opposed to isolated inflammatory events over time. This distinction is particularly true in the case of a first event suggestive of an MS relapse, which may herald a relapsing-remitting (RR) form of MS (85% of MS cases (Jakimovski et al. 2024)). Nevertheless, at age 30, one-third of patients who have experienced such a relapse will remain without new signs of clinical or radiological activity and therefore without a diagnosis of MS (Chung et al. 2020). This necessary diagnostic caution after a first event suggestive of the disease can unfortunately lead to a considerable delay in the start of basic treatment for the disease. If treatment is only started after a new clinical or radiological event, several years may pass, and this might lead to higher levels of disability in the long term (Iaffaldano et al. 2021). The diagnostic criteria for MS, particularly relapsing-remitting MS, have always been designed to speed up diagnosis after the first symptoms suggestive of the disease(McDonald et al. 2001; Polman et al. 2005; 2011; Cohen 2017). This was the case with the first version of the McDonald criteria published in 2001 compared to the previous Poser criteria (McDonald et al. 2001), and subsequent revisions of these criteria have shown a higher percentage of patients diagnosed in the early stages of care compared to this first version (Schwenkenbecher et al. 2018; Hacohen et al. 2020; Gaetani et al. 2018). At the 40th Congress of the European Committee for Treatment and Research in Multiple Sclerosis (ECTRIMS) (September 2024, Copenhagen), the "McDonald 2024" criteria for the diagnosis of MS were presented. They mark a real change from previous versions of these criteria in several respects: * The distinction between relapsing-remitting and progressive forms is no longer central, as the criteria are fairly similar regardless of the mode of onset * Patients with radiologically isolated syndromes can potentially be diagnosed with MS even if they are "presymptomatic" * The possibility of confirming spatial and temporal dissemination using clinical data regardless of additional tests is no longer valid. The definition of spatial and temporal dissemination, and all additional criteria, is based on the results of additiona biological or morphologicall tests. * Spatial dissemination is no longer necessary for diagnosis, which can be made with a single lesion in a typical MRI location (and sufficient additional criteria). * Temporal dissemination is no longer necessary for diagnosis; it becomes an additional criterion like others. * Among the additional criteria that may be required in certain diagnostic situations are the presence of lesions with central vein signs in susceptibility-weighted imaging (SWI) sequences on brain MRI, or the presence of at least one lesion with a paramagnetic rim on the same sequence * The typical locations that are necessarily affected include a fifth location: the optic nerves, whose lesions can be confirmed by ophthalmological examination, visual evoked potentials, or MRI of the optic nerves * Intrathecal synthesis of immunoglobulin G (IgG) can be confirmed by the presence of oligoclonal bands of IgG (OB) in the cderebro spinal fluid (CSF), but also, with these new criteria, by an increase in the kappa light chains index (kappa index). This quantitative method is easily achievable by biological teams with little specialization in the study of OBs and is not expensive. In the literature, an increase in the kappa index is equivalent to the presence of OBs for the diagnosis of MS after a first demyelinating event (Levraut et al. 2023). The calculation of this index could therefore replace the search for OBs, and the new criteria include the possibility of having either test positive as a diagnostic argument. However, the threshold used to consider this index "increased," i.e., in favor of diagnosis, remains to be determined (Hegen et al. 2023). The threshold of 6.1 seems the most likely to be adopted. All these changes still need to be studied in real life: the performance of all these new criteria in terms of the percentage of patients receiving a diagnosis of MS after initial neurological management compared to the previous version (known as "McDonald 2017") is to be determined. "Real life" means: on patients consulting a private or hospital neurologist for an event suggestive of MS and for whom the full set of tests to diagnose MS according to the 2017 and 2024 criteria is used, within the timeframes of routine care (potential delays of several weeks before MRI, before ophthalmological consultation, before lumbar puncture and analysis of IgG and kappa light chains). A previous work studied the performance of the 2017 McDonald criteria in comparison with the three previous versions based on tests performed on 2,220 patients in Lorraine who had experienced a first suspected relapsing-remitting MS event between 2002 and 2020. At the initial assessment (first series of tests and first neurological consultation): 54% of patients had relapsing-remitting MS according to McDonald 2017, compared to 37% according to the previous 2010 version. This 17% increase in the number of patients diagnosed at this stage of care contrasts with the 24-39% increase in patients diagnosed using McDonald 2017 compared to 2010 reported in the existing literature, mainly due to the arbitrary conditions for overselecting patients in the cohorts in the literature and the quasi-experimental use of all available examinations (particularly CSF analysis) (Ferrand et al. 2023). The goal of this study is to document the percentage of patients diagnosed with MS at the start of neurological care using the McDonald 2024 criteria compared to the previous McDonald 2017 criteria after a first event suggestive of MS. This event may have been suggestive of a relapse, progression, or examination showing suggestive lesions without related symptoms, in accordance with the scope of these new criteria. The tests performed to make the diagnosis will include all the techniques proposed by the 2024 criteria, in particular the possibility of using SWI brain imaging data, MRI of the optic nerves and ophthalmological consultation, and the study of kappa light chains in CSF and serum. The deadlines and conditions for performing and analyzing these examinations will be those of routine care, as will the frequency of clinical neurological assessments and patient follow-up. This work will also provide an opportunity to search for an optimal threshold for kappa index positivity in CSF. Primary objective Compare the percentage of patients diagnosed with MS following initial assessment according to the McDonald 2024 criteria and those diagnosed according to the McDonald 2017 criteria among patients seen for suspected MS. Secondary objectives 1. Compare the percentage of patients diagnosed with MS during follow-up according to the McDonald 2024 criteria and those according to the McDonald 2017 criteria. 2. Document the need to use the new examination techniques appearing in the 2024 criteria to make the diagnosis, at the initial stage or during follow-up (brain MRI with SWI, optic nerve MRI, ophthalmological examination data, kappa index). 3. Determine the optimal kappa index threshold to distinguish between patients with OBs and those without (OBs being considered the gold standard) for patients benefiting from this assay in the biochemistry laboratory of the Nancy University Hospital: Definitions Baseline consultation: initial consultation/hospitalization during which the neurologist's opinion regarding the diagnosis of multiple sclerosis is sought • Regardless of the symptoms leading to this consultation: recent symptoms, symptoms dating back several years, several previous episodes rather than just one, isolated radiological findings raising the question of presymptomatic MS, etc. Initial care: period including the baseline consultation and any tests contributing to the suspicion of MS, whether they were performed before the consultation, regardless of how long ago (an MRI scan dating back several years and showing lesions that can be used for diagnosis is considered part of this initial care), or after the consultation, up to a maximum of three months later (any examination after baseline + 90 days is considered a follow-up examination). Tests for suspected MS: any examination that may suggest or contribute to the diagnosis of MS, including: * MRI of the brain +/- optic nerves, regardless of the sequences used. The sequences likely to contribute to the diagnosis of MS are: all T2 and/or FLAIR sequences covering the brain (2D or 3D), all T1 sequences after injection of gadolinium-based contrast agent covering the brain (2D or 3D, spin echo or gradient echo), all magnetic susceptibility sequences covering the brain (SWI, SWAN, etc.), all T2 sequences +/- FLAIR, FSEIR, etc. covering the optic nerves in coronal view, all T1 sequences after injection of gadolinium contrast agent +/- fat signal saturation covering the optic nerves in coronal view. o An examination considered "sufficient brain MRI" contains at least one T2 +/- FLAIR sequence covering the entire brain * An examination considered "sufficient optic nerve MRI" contains at least one T2 +/- FSEIR... coronal sequence on the optic nerves * Spinal cord MRI, including: entire spinal cord MRI and MRI covering only part of the spinal cord: cervical spinal cord MRI (including brain MRI satisfactorily covering the entire cervical spinal cord), thoracolumbar or lumbosacral MRI. The sequences likely to contribute to the diagnosis of MS are: T2 +/- STIR sequences at least sagittal (+/- axial slices) and T1 sequences after injection of gadolinium contrast medium at least sagittal (+/- axial slices). o An examination considered to be "sufficient entire spinal cord MRI" contains at least one sagittal T2 +/- STIR sequence covering the entire cervical-dorsal-lumbar spinal cord (pan-medullary) * CSF study: the two tests useful for diagnosing MS are the search for oligoclonal bands of IgG (isoelectrofocusing of IgG, serum, and CSF) and the calculation of the kappa index (kappa light chains of Ig and albumin, serum, and CSF). A CSF study in which neither of these two tests is performed is not taken into account for this study. o The kappa index will be calculated a priori for patients seen at the Nancy University Hospital only. The positivity threshold for meeting the "CSF" criterion of the 2024 criteria selected for this study is 6.1 ( used in the publication of the criteria). o In a second step, a threshold that may be more optimal than 6.1 will be sought, a threshold that discriminates between patients with OBs and those without. (all other criteria remaining equal). * Ophthalmological consultation: several tests may be performed during this consultation that can contribute to the diagnosis of MS. A consultation in which none of these tests are performed is not included in this study. These tests are: o Measurement of distance visual acuity (significant decrease in visual acuity in one or both eyes) * Fundus examination (edema or atrophy of one or both optic discs) * Measurement of RNFL thickness (or other layers deemed relevant) peripapillary in OCT (increase or decrease in the thickness of one or both RNFLs or other layers deemed relevant) * Visual field testing (scotoma suggestive of optic neuritis) * Visual evoked potential testing (pathological results suggestive of optic neuritis, such as prolongation and/or decreased amplitude of P100) Diagnostic criteria: established using the published criteria and their practical interpretation, as set out in Appendices 1 and 2. • Appendix 1: McDonald 2017 diagnostic criteria for relapsing-remitting MS and primary progressive MS: the two possible forms of onset with these criteria. Patients who did not present any symptoms suggestive of MS at initial assessment, and who were only assessed because an examination (mainly MRI) performed for reasons other than symptoms suggestive of MS revealed one or more lesions suggestive of MS (cases of radiologically isolated syndrome), can never be diagnosed with MS using these 2017 criteria. • Appendix 2: McDonald 2024 diagnostic criteria for MS: all forms of onset (relapsing, progressive, MS-suggestive lesion on MRI without MS-suggestive symptoms). o Please note that the 2024 criteria introduce nuances for minor patients (correct differential diagnosis with ADEM, MOGAD, etc.) that cannot be easily taken into account in this study, and at least one additional criterion is required for patients aged 50 or over at the time of diagnosis • Several cases are possible on the date these criteria are applied to a given patient (at the end of initial treatment for the primary objective): * McDo 17+ McDo24+ * McDo17- McDo24+ * McDo17- McDo24- * McDo17+ McDo24- Methodology Information is provided to the patient during the baseline consultation, based on the principle of non-opposition (research outside the RIPH, since there is no change in standard care). As part of standard care, the neurologist uses all the diagnostic tools necessary to establish a diagnosis of MS according to the current criteria (McDonald 2024). In particular: • A brain MRI with at least one T2 +/- FLAIR 2D or 3D sequence covering the entire brain is considered routine (the absence of at least one brain MRI in the initial management leads to the exclusion of the patient). This MRI scan may have been performed at any time before the baseline consultation and up to 90 days after. * Many of the 2017 and 2024 criteria require additional techniques beyond the T2 +/- FLAIR sequence: T1 sequence with gadolinium injection, SWI sequence, T2 +/- FLAIR and T1 MRI of the optic nerves with gadolinium injection, T2 and T1 MRI of the spinal cord with gadolinium injection, essentially. * The neurologist may request that the initial brain MRI be supplemented with these sequences, in radiology at the CHRU or CHR if these sequences have not been performed (or are uninterpretable) and the diagnosis cannot be made with the MRI alone. The delay must then be a maximum of 90 days after the baseline consultation in order to comply with this "initial care" period. Cerebrospinal fluid (CSF) testing for oligoclonal IgG bands (and the kappa index for patients whose CSF is sent to the biochemistry laboratory at Nancy University Hospital) is routinely performed within 90 days of the baseline consultation (or potentially performed before this baseline visit) (unless there are contraindications for the procedure, the patient refuses, or all criteria are met without the need for this examination). • An ophthalmological consultation will also be performed routinely at the CHRU or CHR (or in private practice if a report is available and with all means implemented to diagnose optic nerve damage) if the diagnosis is not made after the initial examinations (particularly brain MRI), within a maximum of 90 days after the baseline consultation. Each case is reported to the investigating neurologist on an ongoing basis by the neurologist responsible for the baseline consultation at the Nancy University Hospital and Mercy Hospital. A CRF is used for each patient and completed retrospectively by reading the report from the neurologist at Nancy University Hospital or Metz-Thionville Regional Hospital, and using all available sources: MRI reports, ophthalmological consultation reports, MRI image readings, analysis of cerebrospinal fluid test results, etc. The frequency of clinical visits and additional follow-up examinations, the nature of these investigations, and the possible use of treatments, particularly long-term MS treatments, are systematically those of standard care. Only the analysis of the kappa index in the CSF is added for patients whose CSF is sent to the biochemistry laboratory at the Nancy University Hospital (patients seen in consultation at the Nancy University Hospital). This additional analysis does not involve any additional procedures/interventions to those already performed in routine care for the patient. This kappa index analysis will be performed for any CSF and plasma samples sufficient for conventional electrophoresis by isoelectric focusing of immunoglobulin G in CSF and plasma (a reference test already performed routinely for these patients) AND kappa index analysis. The reagent kits will be provided by the industrial laboratory marketing this test, which will be acknowledged in the publication. All patients will be assessed using the 2017 and 2024 criteria by the same neurologist-investigator. Statistical analysis The socio-demographic, clinical, biological, and imaging characteristics of the patients included will be described by their mean and variance for quantitative variables and their proportion for categorical variables. The algorithms for each diagnostic criterion (2017 and 2024) will be applied, and the proportions of diagnostic patients among all subjects included will be calculated. The proportions of subjects diagnosed by each criteria algorithm (presented with their 95% confidence interval) will be compared using a paired Chi² test (McNemar). This comparison will be performed at the initial assessment, at 6 months, one year, and two years of follow-up. The contribution of each symptom (clinical, biological, and imaging) to the establishment of the diagnosis by each algorithm at the initial assessment will be analyzed by calculating their sensitivity and specificity, with their 95% confidence interval, and by calculating the Youden index and accuracy. The search for a discriminating threshold for the light chain kappa index will use various thresholds proposed in the literature, and may also be based on thresholds specific to the sample after visualization of their distribution in the available sample according to the presence or absence of BOC taken as the gold standard. For each threshold, sensitivity and specificity will be calculated. In addition, a ROC curve will be established, with the area under the curve and its 95% confidence interval. The optimal threshold can be chosen based on this curve. Provide for a sensitivity analysis by not using the possibility of confirming spatial and temporal dissemination by the clinic in the 2017 criteria (as these possibilities of confirming spatial and temporal dissemination by the clinic no longer exist in the 2024 criteria, and these criteria in 2017 were a source of significant inaccuracies). Expected degree of statistical significance: P \< 0.05, no correction of the alpha risk threshold planned. Ethics and authorizations Non-interventional research reusing health data acquired in routine care (research not involving human subjects). This research is registered in the data processing registry under MR004 (méthodologie de référence, refenrence methodology) (CNIL : comission nationale de l'informatique et des libertés, National Commission for Information Technology and Civil Liberties, France) under number 2025PI019-663. According to the MR004, the inclusion of patients is based on the principle of non-opposition to the use of their health data for research purposes with the provision of an information document and a form for potential objection to each patient included. Expected results and alternative strategies? A modest improvement in the 2024 diagnostic criteria compared to those of 2017 in terms of the time between the first neurological consultation and the actual diagnosis made by the neurologist. This superiority will probably be more significant in certain specific populations, simply because the 2017 criteria did not really apply to these patients: patients with radiologically isolated syndrome, minor populations, etc. Benefits for patients in the short and medium term. Identifying the real scope of the 2024 criteria, which are more demanding in terms of diagnostic tests, will make it possible to limit the use of these tests to specific cases. This study will also make it possible to refine the real value of these criteria in real life, to guide future versions of these criteria towards a possible simplification of the decision-making algorithms and tests used. If diagnostic "accuracy" is maintained with simplified procedures, the risk of error or misdiagnosis at the crucial moment of diagnosis will be reduced.