|Home | About | Journals | Submit | Contact Us | Français|
Diagnosis of multiple sclerosis (MS) requires exclusion of diseases that could better explain the clinical and paraclinical findings. A systematic process for exclusion of alternative diagnoses has not been defined. An International Panel of MS experts developed consensus perspectives on MS differential diagnosis.
Using available literature and consensus, we developed guidelines for MS differential diagnosis, focusing on exclusion of potential MS mimics, diagnosis of common initial isolated clinical syndromes, and differentiating between MS and non-MS idiopathic inflammatory demyelinating diseases.
We present recommendations for 1) clinical and paraclinical red flags suggesting alternative diagnoses to MS; 2) more precise definition of “clinically isolated syndromes” (CIS), often the first presentations of MS or its alternatives; 3) algorithms for diagnosis of three common CISs related to MS in the optic nerves, brainstem, and spinal cord; and 4) a classification scheme and diagnosis criteria for idiopathic inflammatory demyelinating disorders of the central nervous system.
Differential diagnosis leading to MS or alternatives is complex and a strong evidence base is lacking. Consensus-determined guidelines provide a practical path for diagnosis and will be useful for the non-MS specialist neurologist. Recommendations are made for future research to validate and support these guidelines. Guidance on the differential diagnosis process when MS is under consideration will enhance diagnostic accuracy and precision.
Diagnostic criteria for multiple sclerosis (MS) have evolved over the past 50 years. Although successive versions have differed in emphasis, all have required dissemination of disease in space and time documented by either clinical, paraclinical, or laboratory criteria. Additionally, MS diagnostic criteria have emphasized that alternative explanation for the clinical presentation must be considered and excluded before a diagnosis of MS can be made [1–4].
The most recent McDonald Criteria formally integrate data from magnetic resonance imaging (MRI) and focus on early diagnosis of patients presenting with a clinically isolated syndrome (CIS) suggestive of MS (e.g., unilateral optic neuritis, internuclear ophthalmoplegia, partial myelopathy) [3,4]. Because most such patients will develop a second event over months or years, these diagnostic criteria have been perceived as being prognostic for subsequent disease activity (whether a further relapse will occur), rather than diagnostic (an instrument to differentiate MS from other diseases).
Patients suspected of having MS may have neurological syndromes upon initial examination that are clinically monofocal (no dissemination in space, for which a single CNS lesion can explain signs and symptoms), multifocal (dissemination in space, for which symptoms and signs can only be explained by at least two lesions in separate parts of the CNS) and that have been, over time, monophasic (a single occurrence), multiphasic (relapsing), or progressive in nature. Similar presentations can occur in patients who have an infectious, neoplastic, congenital, metabolic or vascular disease, or non-MS idiopathic inflammatory demyelinating disease (IIDD). Other IIDDs have symptoms that can be similar to those seen in MS (for instance, neuromyelitis optica [NMO], opticospinal MS in Asian populations [OSMS], acute disseminated encephalomyelitis [ADEM]), but differ in course, pathophysiology, treatment, and prognosis (see Figure 1). The ability to make an accurate diagnosis as early as possible is important for patient management, counseling, and optimal therapy.
A conceptual framework for differential diagnosis of MS does not exist. The European MAGNIMS group defined MRI red flags in the setting of clinically suspected MS, which suggest that an alternative diagnosis to MS is likely  but not in the context of other relevant clinical or laboratory examinations and findings. The current article describes an international consensus-based effort to guide the clinical, laboratory, and imaging assessment of patients with a possible diagnosis of MS, so as to help satisfy the requirement for “no better explanation” that is an integral component of all MS diagnostic criteria.
In 2006, the International Advisory Committee on Clinical Trials in MS of the US National MS Society convened the Task Force on Differential Diagnosis in MS to develop a perspective that clinicians may use to address the principle of “no better explanation” for a suspected MS clinical presentation. The Task Force consisted of 18 international (United States, Canada, Europe, Japan) experts in the field of demyelinating disease with differing clinical and research expertise (neurology, ophthalmology, infectious disease, MRI).
The group's initial mission statement was: to provide a data-driven and consensus-based diagnostic approach for patients who present with symptoms and objective clinical evidence suggesting CNS white matter disease; to include guidance for appropriate clinical, radiological, and/or laboratory tests that should be done to exclude alternative diagnoses, especially those that are amenable to appropriate treatment; and to develop a practical tool for neurologists to facilitate accurate diagnosis and to guide management and which will complement the McDonald Diagnostic Criteria. The intent was not to present a comprehensive literature or conceptual review because the spectrum of differential diagnoses is enormous. We focused on patients who present with objective signs that are suggestive of CNS white matter disease and also considered apparently asymptomatic individuals or patients with other common clinically distinct neurological conditions (e.g., migraine) in whom lesions suggestive of white matter disease are showed through MRI.
To address these questions the Task Force formed into three working groups focused on exclusion of potential alternatives to an MS diagnosis, on diagnosis of common initial isolated syndromes when MS is in question, and on differentiating between MS and non-MS IIDDs. Consensus within subgroups and in the Task Force as a whole was reached and recommendations developed through a series of conference calls and subgroup meetings over a year and a meeting of the full Task Force in February 2007. Although informed by published evidence identified by Task Force members when available, the effort did not include a formal literature survey. Opinion-based consensus among the convened experts was developed for diagnostic approaches and classifications for which evidence-based data were lacking.
One subgroup focused on the exclusion of diagnostic alternatives to MS and generated a series of clinical and paraclinical “red flags” that are likely to point away from an MS diagnosis. The group reviewed selected literature relating to demographics, general clinical and neurological findings, paraclinical and laboratory findings (including various imaging techniques and laboratory tests including serum and cerebrospinal fluid [CSF] analyses and evoked potentials) of a spectrum of diseases that might be reasonably considered in the differential diagnoses for MS.
Recognizing that when confronted with a patient with suspected CNS white matter disease for diagnostic evaluation, the first step is to do a clinical exam and order imaging and other laboratory tests to help assess the condition, a table of 79 demographic, clinical, laboratory, and imaging features was prepared. The table was rated independently by the six subgroup members on a 1–5 scale to classify these characteristics as major red flags (rating 4 or 5) that point fairly definitively to a specific non-MS alternative diagnosis or as minor red flags (rating 1 or 2), which suggest that a disorder other than MS should be considered. An intermediate score (rating 3) indicated uncertainty. Scores on each finding from each rater were summed and standard deviations (SD) were calculated for each item (a high SD represents a low degree of concordance among raters). The potential 79 red flags were then classified into three groups according to the following criteria:
A second subgroup focused on CISs that are frequently seen as a first presentation of disease that is eventually diagnosed as MS. The subgroup concluded that the term CIS is confusing in a diagnostic context because it is unclear if it refers to syndromes isolated in time, space, or both and because it lacks pathological specificity [6–9]. A more clear definition of CIS was developed. In addition, the subgroup developed diagnostic algorithms for three of the most typical CISs (optic neuropathy, brain stem, and spinal cord syndromes) and distinguished between features of CIS that commonly precede MS versus uncommon or atypical features that merit consideration of alternative diagnoses and an expanded evaluation.
A third subgroup evaluated clinical, demographic, and paraclinical factors that differentiate prototypic MS from “variants,” such as NMO and ADEM, and proposed consensus criteria for their diagnosis in light of recent data on specific imaging features and biomarkers. The subgroup also developed a working classification of IIDDs, recognizing that data on which to base such a classification was scant.
The Panel agreed that in making a differential diagnosis, patients with presentations suggestive of MS should be evaluated using a sequential strategy and presented its findings accordingly:
Differential diagnostic schemes have devoted little attention to patients with clinical presentations of CNS disease similar to MS but which may not develop into MS. Such patients may include those who eventually are determined to have, for example, vascular or infectious disorders. Diagnostic evaluation strategies apply to individuals with:
Table 1 presents 79 clinical and paraclinical red flag findings of patients who present with CNS disease for which MS is being considered. Through a ballot process among the subgroup members, as described in the “Methods” section, 36 major red flags were identified that point fairly definitively to a non-MS alternative diagnosis, the majority of which are clinical in nature. Eleven minor red flags were identified that suggest while MS is a possible diagnosis, a disorder other than MS should be considered, and that a decision cannot be made simply based on the noted assessment alone. An additional 32 clinical, paraclinical, and laboratory assays, many of which are imaging findings, were determined to be of intermediate weight. These received a relatively high mean score for weighting in the consensus process, but with high SDs among the raters, indicating a lack of consensus on their weight and importance. Their value for differential diagnosis should be considered in the overall context in which they appear (i.e., additional informative clinical, laboratory, or paraclinical findings).
At first presentation when MS is in question, there may already be a history of more than one relapse or of a progressive course from onset. However, most patients eventually diagnosed with MS present with an acute clinical episode reflecting CNS white matter pathology, generally called a CIS. Because this occurs frequently and because the absence of information indicating dissemination in time precludes an immediate diagnosis of MS, we focused on differential diagnosis of such isolated syndromes.
CIS presentations most commonly involve a single optic nerve, the spinal cord, or the brain stem, although other isolated syndromes may occur, such as ones affecting the cerebral hemispheres (e.g., hemianopia). Although widely used to signal the first presentation of demyelinating disease, the term CIS is inadequate: it has been variably used to group patients with: i) a single clinical event and signs that indicate a single lesion only (thus disease isolated in space and time); ii) recurrent episodes in a single location (thus with disease isolated in space but not in time, e.g. recurrent optic neuritis) ; and iii) a single clinical event where the symptoms and/or neurological examination findings suggest the presence of two or more lesions in separate locations (thus isolated in time but not in space).
Most clinical trials of patients with CIS aimed at determining the role of interferon-β in delaying time to clinically definite MS included patients whose clinical episode was monophasic but who were polysymptomatic and whose examination showed multifocal CNS signs (e.g., a patient with acute optic neuritis and Lhermitte's sign or a patient with optic neuritis who also has an extensor plantar response) [7,8]. In one CIS trial, 48% of patients with CIS had evidence for multifocal disease . Strategies have been proposed to reduce variability in interpreting clinical trial outcomes by distinguishing initial monofocal from multifocal presentations and stratifying enrolment accordingly . Furthermore, the term CIS ignores first presentations that may not be clinical but may be detected by paraclinical and laboratory findings , and does not discriminate between patients who have a single clinical presentation with or without additional symptomatic lesions on MRI, two entities that have different prognoses [13–16].
Because of these ambiguities of interpretation and the importance of CIS in the differential diagnosis process, the Panel felt it important to more closely define the term. We agreed that a CIS should be defined as a monophasic presentation with suspected underlying inflammatory demyelinating disease. “Monophasic presentation” implies a single clinical episode at first presentation that is of relatively rapid onset. Multiple simultaneous clinical/ paraclinical presentations (representing dissemination in space) are possible, although dissemination in time should not be evident. Four classes of CIS can thus be defined based on whether the monophasic clinical presentation has mono- or multifocal clinical or MRI features (Table 2; CIS types 1–4). Finally, there should be reasonable grounds for suspecting inflammatory demyelinating disease as the underlying pathology.
MRI can influence a decision as to whether a presentation is due to monofocal or multifocal lesions and the likelihood for an ultimate diagnosis of MS (Table 2). Patients with at least one asymptomatic MRI lesion characteristic of demyelination have a high probability of later meeting criteria for MS (CIS Types 1 and 2); the prognosis varies but is not strongly based on the number and location of lesions [13,14]. Patients with monofocal clinical presentations and no asymptomatic lesions characteristic of demyelination have a relatively low risk for later meeting criteria for MS (CIS Type 3) . A multifocal clinical presentation without MRI-detected asymptomatic lesions is likely rare (CIS Type 4) and such patients require further follow-up to determine if they have MS or another condition.
Although symptoms and signs of a monophasic illness have been an essential prerequisite for diagnosis of CIS, there is an exceptional scenario that the Panel feels warrants inclusion as CIS Type 5 (Table 2): patients who have no symptoms or only non-specific symptoms (e.g., headache, dizziness), but have MRI evidence for multifocal abnormalities typical for demyelination. Such patients are increasingly identified using MRI for incidental indications (e.g., headache) especially with high-field strength magnets with greater sensitivity for such lesions . Current criteria preclude a diagnosis of MS without objective clinical evidence for CNS abnormality and the ability to establish a confident diagnosis of MS in such individuals and their natural history should be addressed through prospective studies .
A wide array of isolated CNS syndromes encountered in a diagnostic work-up for possible MS can be included in the CIS definition. As noted, some are more strongly predictive of an eventual MS diagnosis than others, depending on imaging characteristics but their clinical appearance at initial presentation can also provide clues to the likelihood of an eventual MS diagnosis. Table 3 categorizes CIS presentations of patients eventually diagnosed as having MS into i) those that are typical of patients eventually diagnosed as having MS; (ii) those that are less common but nevertheless may be an initial presentation in patients eventually diagnosed with MS, or may signal another disease; and (iii) those that are atypical and suggest an alternative diagnosis.
Perhaps the three most common CIS syndromes seen at presentation in the MS diagnosis process include those affecting the optic nerve, brain stem, and spinal cord. Flow diagrams illustrating an approach to their differential diagnosis are presented in Figures 2–4. The flow diagrams are presented to illustrate some principles of the clinical and laboratory evaluation and do not aim to be comprehensive; they emphasize typical and readily available investigations and outcomes.
Although MS may be the most frequent ultimate diagnosis of a case presenting with IIDD, clinical, radiological, and immunological biomarkers that persist over time may help distinguish and define subgroups of IIDD from MS. We focused on two of the most common differential diagnoses for IIDD: NMO and ADEM. The Panel built upon recently proposed criteria for NMO that are 90% sensitive and specific in differentiating NMO from MS [19,20]. However, criteria for diagnosis of ADEM are not well validated and about 30% of patients who meet the ADEM criteria at initial presentation will later receive a diagnosis of MS [21–24]. The rate of conversion from ADEM to MS will be higher in adults (in whom ADEM is less common) than in children, in whom about 20% of those initially diagnosed with ADEM will ultimately be diagnosed with MS.
NMO is perhaps the most commonly seen non-MS IIDD [25,26]. It has been distinguished in the past from MS because of largely restricted manifestations of optic neuritis and myelitis and by its monophasic, not relapsing, course. However, more contemporary studies suggest that NMO is usually a relapsing IIDD, blurring the distinction between NMO and MS [27–31]. Normal brain imaging and longitudinally extensive cord lesions in the context of acute myelitis have helped to distinguish NMO from MS. A recently discovered, highly specific, and moderately sensitive serum biomarker, NMO-IgG, is useful for diagnosis of NMO [32,33].
In distinguishing NMO from MS, the subgroup agreed that:
Some subjects presenting with IIDD have recurrent transverse myelitis alone accompanied by long spinal cord lesions or recurrent optic neuritis alone and are seropositive for NMO-IgG. These may represent limited or inaugural syndromes of NMO. Although many clinicians believe that such patients should be treated as if they have NMO [26,44,45], until the relationship between isolated recurrent transverse myelitis or isolated recurrent optic neuritis and NMO is better established, the Panel concluded that these spatially limited syndromes should not qualify as NMO, even in the presence of a positive NMO-IgG serum assay. The subsequent development of optic neuritis in a patient with myelitis or vice versa may permit a later diagnosis of NMO.
Biopsy evidence of sarcoidosis or vasculitis, which can occasionally cause optic neuritis and myelitis, excludes NMO. Patients with optic neuritis, myelitis, or both occurring with preexisting or subsequently developing systemic lupus erythema-tosus or Sjögren's syndrome, a not infrequent clinical occurrence, are seropositive for NMO-IgG at the same frequency as in “uncomplicated NMO” and may have both NMO and organ-specific or non-organ specific autoimmunity . However, the Panel concluded (conservatively) that pending further study, clinical evidence of systemic lupus erythematosus or Sjögren's should exclude a diagnosis of NMO. Seropositivity for antinuclear antibodies (ANA) or Sjögren's (SSA/SSB), which is commonly uncovered in the evaluation of patients for NMO, does not exclude the diagnosis of NMO when clinical evidence for systemic lupus erythe-matosus or Sjögren's syndrome is lacking.
Asian optic-spinal forms of MS (OSMS) may be confused with NMO. It is unclear whether differences between OSMS and NMO in Asia and Western countries are due to biological differences or to differences in nomenclature. In Asia, patients with optic neuritis and myelitis, regardless of the length of their spinal cord lesion, are classified as OSMS, whereas such patients without LETM lesions in most instances would be classified as having typical MS in Western countries. Furthermore, whenever clinical or radiological involvement of the brain is detected (except when confined to the brainstem or hypothalamus), MS is diagnosed in Asia [47,48], whereas such patients who fulfill other criteria for NMO in Western countries are typically diagnosed with NMO. Continued natural history studies of such patients will be necessary to determine if clinical predictions about their phenotypes and responses to therapy are adequately predicted by the diagnosis assigned using Asian and Western approaches.
ADEM has been historically recognized as distinct from MS based on its monophasic course and encephalopathy or coma in combination with mul-tifocal symptoms (e.g., cerebellar signs, cerebral motor or sensory features, optic neuritis or myelitis) characteristic of an IIDD, often following an infectious illness. MRI typically shows usually symmetrical multifocal or diffuse brain lesions . Even when conservatively defined by requiring encepha-lopathy, an initial diagnosis of ADEM is often revised to prototypic MS after evidence emerges for continuing clinical activity consistent with MS . It recently has been argued that the traditional requirement of a monophasic course for ADEM might be too strict and that some patients experience recurrence of their initial ADEM symptoms with re-emergence of the same MRI lesions as were present at the time of their initial illness (recurrent ADEM) [21–24]. Although characteristics such as encephalopathy with multifocal symptoms typical of IIDD may make ADEM more likely than MS, no clinical, paraclinical, or imaging criteria reliably distinguish fulminant initial episodes of MS from ADEM (Table 5) .
Recently proposed definitions for pediatric MS and ADEM were designed to minimize overlap between the two disorders, by requiring encepha-lopathy as an essential component of the ADEM definition . Although these criteria are consensus-driven and based on pediatric cohorts, it is reasonable to propose their use in adult populations as well, pending further evaluation of differences between ADEM in children and adults. These criteria emphasize the relative specificity of encephalopathy in ADEM although they do not distinguish between degrees of encephalopathy, which can vary from irritability to coma; severe encepha-lopathy may be more specific for ADEM than milder encephalopathy.
The Panel proposed that a diagnosis of ADEM should be made in patients with a first event compatible with demyelinating disease that is acute or sub-acute in onset (over days to weeks), with a stable or stuttering course, but only when additional characteristics are present. Encephalopathy, manifest either as altered level of consciousness, behavioral change, or altered cognitive function, should be present. New symptoms may emerge over intervals up to 3 months from onset without intervening remission (but not beyond 3 months). However, if a remission of the initial symptoms occurs followed by new symptoms after an interval of 1 month, MS is more likely than ADEM. Although MRI is non-specific in ADEM, the presence of multiple supra- or infra-tentorial lesions in combination with lesions of deep grey nuclei and at least one lesion greater than 1–2 cm in diameter are characteristic. Spinal cord lesions may or may not be present, but when present, tend to be longitudinally extensive (Table 5).
Very rare ADEM patients may experience recurrence of initial symptoms and signs after 3 months without development of new lesions although existing lesions may enlarge or re-enhance with gadolinium [52,53], in other words a recurrent
course as distinct from a multiphasic course which might include new symptoms and new lesions. If no other diagnosis is apparent after appropriate evaluation, the diagnosis “recurrent ADEM” is warranted. However, the emergence of new lesions with different symptoms beyond a 3-month interval from onset likely signals MS regardless of the specific clinical characteristics. The Panel did not endorse a diagnosis of “multiphasic ADEM” when new lesions and different symptoms emerge over time [52,53] because of the inability to distinguish such presentations from MS and because such a diagnosis could result in a delay in use of MS-specific immunotherapy. The Panel concluded that most such patients will continue to have inflammatory disease activity that is characteristic of MS.
Recognizing the differences between MS, NMO, ADEM, and their variants, a classification for IIDDs was developed, which should be considered whenever IIDD is the working diagnosis (Table 6). MS is diagnosed when McDonald criteria for dissemination of symptoms/signs in space and time are reported [3,4] and includes relapsing-remitting, secondary progressive, primary progressive, and progressive-relapsing forms . Subjects who present with atypical IIDD syndromes, such as Marburg's variant of MS [55,56], Balo's concentric sclerosis , and other tumefactive forms of demyelinating disease [58,59] are considered to be “unclassifiable” in terms of a possible MS diagnosis at first event, but if and when criteria for dissemination in space and time are satisfied, a diagnosis of MS might be considered appropriate in spite of an atypical initial or subsequent presentation. Alternative diagnoses, such as cerebral lymphoma, glioma-tosis cerebri, or vasculitis should strongly be considered as well. Some multi-episode relapsing presentations, such as relapsing transverse myelitis, may also be unclassifiable in terms of potential for an MS diagnosis at presentation when there is neither clinical nor MRI evidence for dissemination in space. NMO and ADEM are separate conditions in the category of IIDDs and occur in both monophasic and recurrent forms. Some initial presentations of CIS are truly monophasic inflammatory demyelinating conditions, but many patients with CIS syndromes, if not most, will develop MS.
The range of diseases that must be excluded to make a diagnosis of MS is wide. Although the exclusion of more likely alternatives is a critical aspect of MS diagnosis, little attention has been paid to providing guidance for clinicians. The Panel has delineated two broad categories of diseases that should be considered in the differential diagnosis work-up of CNS disease suggestive of MS: non-IIDDs with symptoms and signs that may mimic IIDDs (including MS) and non-MS IIDDs, some of which are also relapsing and remitting. We have suggested a classification of clinical presentations into those that are typical and suggestive for MS and those that are atypical for MS, and have described a series of red flags that might signal a more likely alternative diagnosis. Our conclusions and recommendations are largely consensus driven and should be assessed in prospective clinical studies, preferably in large multicenter efforts that include both specialized MS referral centers and non-specialized clinical settings to test their applicability in both environments. Such studies could establish diagnostic algorithms that provide an accurate diagnosis for CNS syndromes suggestive of MS.
Prospective studies should include all CIS syndromes as redefined here, including those with normal MRI findings and those with asymptomatic lesions highly suggestive of MS that are detected incidentally. A first descriptive retrospective study of patients with incidentally detected asymptomatic lesions indicated that a subset of such patients frequently shows rapid early conversion to early clinically defined MS . Prospective long-term follow-up is needed to assess the natural history of such individuals. About 30% of patients with an otherwise typical isolated CNS syndrome suggestive of demyelination have completely normal MRI apart from the symptomatic lesion(s) [16,40,60]. Ten to 14-year follow-up has shown that about 20% of such patients develop clinically definite MS [13,15]. Although some MRI-negative CIS patients have CSF oligoclonal bands typically seen in patients with MS, others do not . Such patients may have a truly isolated syndrome in space and time and CIS may be the only diagnosis possible .
Both clinical and paraclinical assessments inform the differential diagnosis and evaluation of a patient. We have presented a series of consensus-based red flag criteria that the Panel agrees signal a high likelihood that a disease at initial presentation is not MS. The sensitivity, specificity, and accuracy of these red flags have not been investigated. Each red flag was rated by the Panel in isolation and not in a larger clinical and/paraclinical context; this is rarely the situation confronting a clinician making a diagnosis. The differential diagnosis may be different if a given red flag is the sole finding or appears in conjunction with other relevant symptoms, signs, and paraclinical and laboratory abnormalities. The value and meaning of an isolated red flag can often be resolved in context of demographic, clinical, and paraclinical findings. For instance, onset under age 20 is listed as a minor red flag. However, in the setting of diffuse and symmetrical white matter changes suggestive of leukodystrophy, young onset would be a major red flag suggesting a different diagnosis than MS. However, in the setting of facial numbness or optic neuritis, it is not a red flag at all. Defining sets of symptoms/signs and red flags that would increase the diagnostic confidence in making a diagnosis of MS versus other conditions is a high research priority.
The most common alternative conditions that mimic MS may be the most difficult to eliminate in differential diagnosis. The relative prevalence of disorders in a specific geographic area or population should be considered in reaching a diagnosis. However, the lack of accurate prevalence data for many alternative diagnoses (and their associated red flags) relative to MS currently makes it difficult to weight diagnoses based on prevalence.
Although our discussions included a detailed evaluation of MS in Asian populations and the characteristics of Asian OSMS versus NMO and MS, most data related to the differential diagnosis of IIDD arise from studies conducted in Western populations. The frameworks proposed here may be less applicable to non-Western populations, and this issue should be addressed prospectively. Regional ethnically based differences in disease definition continue to complicate the determination of similarities and differences between NMO and Asian OSMS and confound the interpretation of data on seropositivity rates for NMO-IgG in NMO and in Asian OSMS [32,33,48]. A worldwide consensus definition would facilitate further clinical and biomarker research.
Disease biomarkers will aid enormously in differential diagnosis. Accurate and sensitive disease markers – imaging or laboratory based – may provide non-invasive aids to differential diagnosis. The discovery of NMO-IgG that helps to distinguish NMO from MS is a good example. Relevant imaging advances may include non-conventional MR techniques to quantitate change in normal appearing white matter that may be relatively specific for MS, and high-field MRI to better visualize Dawson's fingers  or cortical lesions that may be specific to MS.
The Panel adopted diagnostic criteria for NMO that were modified from previously presented criteria , which must be considered tentative and subject to further international study and validation. Other groups have reported similar sensitivity and specificity to those originally reported and thereby provide some independent validation . The criteria may require revision pending further investigation into the relationship between symptoms of NMO and systemic autoimmune diseases such as systemic lupus erythematosus . It is unclear if limited presentations of NMO (e.g., recurrent optic neuritis; recurrent transverse myelitis with longitudinally extensive spinal cord lesions) in patients seropositive for aquaporin-4 autoantibodies are NMO spectrum diseases . Although NMO is usually a relapsing disease, it may also be monophasic; detailed long-term follow-up of such cases may provide insights into pathological differences between the two.
The recommended criteria for ADEM (requiring encephalopathy and a limited time frame for recurrent ADEM events) and MS  should facilitate prompt differential diagnosis and consequently different recommendations for counseling and treatment based on the prognosis. Prospective follow-up studies of cases of ADEM are recommended to investigate the performance of these criteria.
This detailed consideration to differential diagnosis at first clinical presentation when MS is being considered should help to refine a key provision of the McDonald diagnostic criteria – the need to exclude other more likely explanations than MS for the clinical presentation. Our recommendations are largely consensus-based and evidence to support many of the recommendations is currently lacking. Algorithms and other recommendations for diagnostic procedures need to be tested prospectively in a relevant spectrum of clinical settings internationally to confirm or refine their utility.
The authors gratefully acknowledge constructive criticism provided in review of an earlier version of the manuscript by Drs Sten Fredrikson (Copenhagen), Andrew Goodman (Rochester, NY), Catherine Lubetzki (Paris), John Newsom-Davis (Oxford, deceased), Paul O'Connor (Toronto), Richard Rudick (Cleveland), and Jack Simon (Portland, OR). The work of the Task Force was supported by the National Multiple Sclerosis Society (US).