MRI of the relevant domain should be performed to confirm whether clinical symptoms represent an attack of NMOSD: Yes

Neuromyelitis optica spectrum disorder (NMOSD) is an autoimmune central nervous system (CNS) disease associated with pathogenic autoantibodies that target aquaporin-4 (AQP4).^1,2 The disease declares that it is active with emergence of acute symptoms that worsen over hours to days before they plateau.^1,2 These clinical events are called attacks (inaugural event) or relapses (subsequent events; the terms are often used interchangeably). Hallmark attack types in NMOSD include optic neuritis, transverse myelitis, and area postrema syndrome, with other CNS regions targeted less commonly.^1,2 Identification of one or more characteristic attack types is necessary to confirm a diagnosis of NMOSD with or without AQP4 antibody (AQP4-IgG). ¹ Current diagnostic criteria integrate magnetic resonance imaging (MRI) lesion patterns that accompany less specific clinical syndromes affecting the cerebrum, diencephalon, or brainstem and are especially useful in AQP4-IgG seronegative cases. ¹

Approximately 90% of patients with NMOSD seropositive for AQP4-IgG experience a relapsing course with periods of disease quiescence and clinical stability occurring between relapses.^1,2 Functional recovery from relapses is variable but usually incomplete. ³ Neurological impairment and disability accrue in stepwise fashion with successive relapses, ³ and a latent secondary progressive clinical course is rare. ¹ Therefore, relapse prevention is the primary therapeutic goal.^1–4 Confident clinical relapse identification confirms new inflammatory disease activity, signaling the need for acute therapy and initiation or change of preventive therapy.^1–4

Despite the importance of relapses, validated and universally accepted criteria for relapse definition are lacking. Relapses are usually defined as acute onset of new or worsening neurologic symptoms attributable to NMOSD, lasting at least 24 hours, occurring after >30 days of preceding clinical stability, and free of confounding factors (e.g. fever, infection, medications effect, among others).^1,5 Severe and typical optic neuritis and myelitis relapses usually appear straightforward to identify owing to specific localizing signs. However, in most clinical scenarios, obtaining additional objective MRI data meaningfully enhances specificity and sensitivity. ⁶ First, consider a presentation of isolated nausea, vomiting, or hiccups; although potentially owing to NMOSD-related area postrema syndrome, these symptoms are non-specific and have many common alternate medical causes. ⁶ Second, less common NMOSD events affecting brain or diencephalon are clinically non-specific and require MRI to detect the signature NMOSD lesion patterns that add necessary specificity and exclude alternative diagnoses. ⁶ Finally, even for the canonical relapse types of myelitis and optic neuritis, supplementary evidence may be needed for relapse confirmation. For example, clinical events that present with sensory or other subjective symptoms and those that are less severe or present for evaluation very early after symptom onset may not be accompanied by new objective neurological signs. Furthermore, there may be uncertainty regarding the source of symptoms when events are confounded by the effects of non-NMOSD comorbidities (e.g. pain, orthopedic disorders, medical conditions) or the degree of clinical change that is attributable to new NMOSD activity when symptoms localize to previously affected CNS regions with pre-existing clinical deficits. ⁷ In these instances, gadolinium-enhanced MRI of the affected neuroanatomic region may provide objective evidence that supports or refutes an NMOSD relapse. In one study, 23% of suspected relapses did not have a new or enhancing MRI lesion correlate in the affected clinical domain, suggesting a diagnosis of pseudorelapse, and distinguishing spinal cord relapse from pseudorelapse was most challenging. ⁸

The recent randomized controlled trials of preventive immunotherapies (eculizumab, inebilizumab, and satralizumab) have sharpened the focus on the approach to relapse confirmation. ⁴ The primary outcome of each trial was time to first relapse confirmed by an expert adjudication committee. ⁴ Each trial utilized similar clinical criteria requiring a >24-hour event associated with NMOSD-related objective clinical change, typically assessed using the Functional Systems components of the Expanded Disability Status Scale. The N-MOmentum trial of inebilizumab ⁵ was the only trial to routinely incorporate MRI into relapse assessment. It employed a set of 18 clinical criteria across the domains of optic neuritis, myelitis, area postrema syndrome, brainstem, or brain. The criteria also integrated MRI findings (new or enlarging T2 lesions and/or new gadolinium-enhancing lesions) when necessary. Review of MRI data was deemed necessary during adjudication of 39% of reported events and 33% of committee-adjudicated relapses. ⁵ Moreover, retrospective assessment showed new domain-specific T1 and T2 lesions in 90% of adjudicated relapses. ⁹ These results require confirmation and validation but, in this setting, where trial results hinge upon dichotomous declarations about relapse occurrence, the quest to identify the true positives and true negatives clearly favors adjunctive MRI use.

The results of the clinical trials and their open-label extensions confirmed the importance of relapse prevention and its association with preserving neurological function. ⁴ The focus on relapse assessment is now heightened in clinical practice, where detection of a true relapse in a patient on preventive therapy leads to examination of medication compliance and consideration of change in therapy while a negative relapse assessment avoids both unnecessary acute intervention and preventive medication change. Evidence suggests that these issues already influence routine practice behavior. An international survey of experts reported that approximately 80% used MRI “always or regularly” to confirm various suspected relapse types. ¹⁰ Overall MRI accessibility was immediate for 59% and within 72 hours for 96.2% of centers. Notably, immediate access to MRI or MRI with contrast was 80% in North American centers compared with 55% (53% with contrast) in Europe and 38% (31% with contrast) in other participating countries. Thus, the use of optimally timed MRI assessment, especially in lower-access countries, may be challenging. However, since MRI access within 72 hours is nearly universal in surveyed countries, relapse definitions that include MRI criteria for clinical practice use appear both desirable and feasible.

Validation studies are needed to confirm and quantify the added value of MRI in relapse assessment, and several practical issues (MRI protocols, timing, quality, access, among others) remain. However, we submit that evidence from clinicoradiological correlations, practice experience, and controlled trial data all strongly support the routine use of adjunctive contrast–enhanced MRI of the relevant clinical domain, where available, for NMOSD relapse confirmation in practice and should be required in all clinical research studies where relapse is a key outcome measure.