Double-Negative Myelitis Represents a Diverse Classification State Rather Than Single Disease

A comprehensive narrative review reveals that double-seronegative acute inflammatory myelopathy represents a provisional serological classification state rather than a distinct biological disease entity. Researchers Ping Fan, Youlin Tan, Zhen Ren, and Qi Zhang synthesized evidence spanning diagnostic frameworks, antibody testing platforms, neuroimaging patterns, and longitudinal follow-up published through August 2026. Acute inflammatory myelopathies constitute a diverse cluster of acute and subacute spinal cord syndromes presenting with clinical features indicative of inflammatory damage. Historically categorized under broad syndromic descriptions such as acute transverse myelitis, these clinical conditions encompass multiple distinct underlying mechanisms, clinical courses, and therapeutic requirements. The synthesis establishes that while the discovery of aquaporin-4 and myelin oligodendrocyte glycoprotein antibodies has driven a profound transition from syndrome-based nomenclature toward disease-specific classification, patients testing negative for both biomarkers encompass an exceptionally heterogeneous population. This double-negative group includes neuroinflammatory disorders and non-inflammatory mimics whose true diagnostic attribution frequently evolves over extended longitudinal observation.

For decades, clinical neurology relied heavily on syndromic descriptions to categorize acute spinal cord inflammation. The diagnostic framework established by the Transverse Myelitis Collaborative Group in 2002 laid a foundational structure for separating idiopathic transverse myelitis from secondary causes. However, acute transverse myelitis primarily reflects an anatomical phenotype of spinal cord involvement rather than a unified etiological diagnosis. Because secondary causes include immune-mediated disorders, systemic infections, vascular disruptions, and neoplastic conditions—each dictating fundamentally different therapeutic strategies and carrying divergent long-term prognoses—syndromic classification proved insufficient for targeted clinical care. The identification of aquaporin-4 immunoglobulin G (AQP4-IgG) initiated a paradigm shift, transforming neuromyelitis optica from a purely clinical syndrome requiring optic neuritis and myelitis into neuromyelitis optica spectrum disorder (NMOSD) defined by a concrete serological biomarker. In 2015, the International Panel for NMO Diagnosis formalized stratified diagnostic pathways, allowing seropositive patients to receive diagnoses following core clinical events while mandating stricter imaging and clinical criteria for seronegative individuals.

Diagnostic refinement accelerated further following the characterization of myelin oligodendrocyte glycoprotein immunoglobulin G (MOG-IgG) and the establishment of international criteria for MOG antibody-associated disease (MOGAD) in 2023. These diagnostic criteria enabled clinicians to reclassify many patients previously designated as AQP4-IgG-negative NMOSD or unclassified central demyelinating syndromes into a distinct disease category. The review highlights that validating MOGAD requires integrating reliable laboratory detection with compatible core clinical demyelinating presentations. The 2023 guidelines specifically emphasize utilizing cell-based assays that present full-length human MOG as the target antigen. Diagnostic interpretation must account for pre-test probability, as low-positive antibody results carry a substantial risk of false attribution in patients lacking prototypical demyelinating presentations. Furthermore, biological titers fluctuate dynamically during the clinical course; delayed serum collection after symptom onset or following acute immunotherapy can diminish antibody titers below detection thresholds, potentially miscategorizing seropositive individuals as double-seronegative.

Despite these serological advances, a significant subset of patients presenting with acute inflammatory myelopathy tests negative for both AQP4-IgG and MOG-IgG. The authors demonstrate that the label “double-negative” actually encompasses two fundamentally distinct clinical groups rather than a uniform cohort. The first subgroup comprises strictly defined double-negative NMOSD, where patients fulfill rigorous 2015 consensus diagnostic criteria for seronegative disease with alternative inflammatory and non-inflammatory causes systematically excluded. The second subgroup consists of individuals provisionally labeled double-negative purely because both antibody assays returned negative results, leaving them without an assigned disease-specific category. While the former represents a criteria-defined clinical diagnosis, the latter constitutes an unassigned serological state. Consequently, interpreting double-seronegative findings requires careful distinction between a finalized clinical diagnosis and an unresolved diagnostic holding category, particularly because clinical composition, relapse rates, and therapeutic responsiveness vary substantially between these two populations across different clinical cohorts.

The review emphasizes that technical differences in antibody testing platforms exert a profound impact on whether patients are classified as seronegative. Serum cell-based assays represent the standard reference methodology for detecting both AQP4-IgG and MOG-IgG, yet analytical sensitivity varies across live-cell and fixed-cell platforms. Multicenter studies show that while cell-based assay platforms exhibit high overall concordance, analytical discrepancies emerge around low-positive samples, with live-cell assays demonstrating higher sensitivity in certain comparative studies. Crucially, negative findings produced by older diagnostic techniques—such as enzyme-linked immunosorbent assays (ELISA) or tissue-based indirect immunofluorescence—cannot be equated with high-quality cell-based assay results. Beyond platform variations, sampling timing and therapeutic interventions directly modulate assay sensitivity. Specifically, immunotherapy and treatment exposure prior to blood sampling can affect antibody detection results. Clinicians must therefore interpret double-seronegative results in the specific context of assay format, sampling latency, and preceding immunotherapies.

Although autoantibody profiling has reorganized neuroimmunology, neuroimaging remains essential for clinical localization and differential classification. The review details distinctive neuroimaging tendencies across the three primary central inflammatory demyelinating diseases: multiple sclerosis, AQP4-IgG-positive NMOSD, and MOGAD. Multiple sclerosis typically presents with short-segment spinal cord lesions occupying peripheral or posterolateral cord regions. Conversely, AQP4-IgG-positive NMOSD is more commonly associated with longitudinally extensive transverse myelitis, central cord involvement, bright spotty lesions, and extensive cross-sectional involvement. MOGAD frequently demonstrates gray matter-predominant lesions, an axial H-sign, and conus medullaris predilection. However, a 2025 systematic review of 66 studies underscored that individual imaging signs exhibit substantial overlap. Longitudinally extensive lesions occur in both NMOSD and MOGAD, and lesion length can change dramatically over time as acute inflammation resolves. Furthermore, initial spinal MRI scans may appear completely normal in a subset of verified MOGAD cases.

Cerebrospinal fluid analysis provides complementary diagnostic evidence when evaluating double-seronegative spinal cord syndromes. Persistent oligoclonal bands, alongside characteristic brain white matter lesions fulfilling the 2024 revised McDonald criteria, strongly favor multiple sclerosis. In contrast, AQP4-IgG-positive NMOSD and MOGAD typically display marked pleocytosis and elevated protein levels during acute attacks, whereas oligoclonal bands are usually absent or transient. Marked CSF abnormalities—such as profoundly decreased glucose, extreme hyperproteinorachia, or unresolving inflammatory pleocytosis—prompt evaluation for conditions beyond classic demyelination, including neurosarcoidosis. Neurosarcoidosis often presents with subacute or chronically progressive myelopathy exhibiting persistent contrast enhancement and leptomeningeal or nerve root involvement, supported by systemic disease manifestations. Another inflammatory entity, glial fibrillary acidic protein (GFAP) astrocytopathy, commonly manifests as meningoencephalitis or encephalomyelitis rather than isolated myelopathy. Diagnosis of GFAP astrocytopathy is supported by encephalopathic symptoms, marked CSF inflammation, linear perivascular radial brain enhancement, and the presence of CSF GFAP-IgG antibodies.

In patients presenting with acute myelopathy alongside systemic autoimmune antibodies, diagnostic attribution requires caution. The presence of antinuclear antibodies, anti-SSA/Ro, or antiphospholipid antibodies demonstrates systemic immune dysregulation but cannot establish the spinal cord lesion’s etiology. Research demonstrates frequent overlap between AQP4-IgG-positive NMOSD and systemic rheumatologic diseases, particularly Sjögren’s disease, emphasizing that rheumatology patients presenting with transverse myelitis require specific AQP4-IgG screening. Furthermore, double-seronegative cohorts frequently capture non-inflammatory vascular and structural mimics that masquerade as inflammatory myelitis. Such cases indicate that some patients entering double-negative cohorts do not have immune-mediated inflammatory myelopathy, making disease course and imaging features essential criteria for determining the underlying etiology.

Longitudinal clinical observation demonstrates that diagnostic attribution in double-seronegative myelopathy is inherently dynamic and time-dependent. In early cohorts initially classified as AQP4-IgG-negative longitudinally extensive transverse myelitis, extended follow-up and subsequent testing reclassified a substantial proportion of patients with MOGAD, multiple sclerosis, or alternative conditions. Among patients who remain double-seronegative, clinical trajectories range from monophasic episodes to severe relapsing disease. The review highlights a 2026 systematic review and meta-analysis by Malvaso and colleagues examining 41 studies and 671 individuals diagnosed with double-negative NMOSD. Within that meta-analysis, 458 of 622 patients with documented course data (73.6 percent) experienced a relapsing disease trajectory. However, the authors emphasize that published recurrence rates are heavily shaped by referral bias. Tertiary referral centers disproportionately enroll severe, treatment-refractory, and recurrent presentations, whereas community cohorts tracking first-time isolated double-negative myelitis observe far higher frequencies of monophasic disease, meaning relapse statistics cannot be generalized universally.

To address current diagnostic uncertainties, the authors outline critical priorities for future neuroimmunological research. Existing literature suffers from definitional drift, where older cohorts classified as seronegative NMOSD were evaluated before MOG-IgG assays were developed or using earlier antibody detection methods with lower analytical sensitivity. Standardizing cell-based testing protocols, validating assay performance across laboratories, and defining consistent diagnostic criteria are essential to harmonize research cohorts. Methodologically, research must maintain clear distinctions among strictly criteria-defined double-negative NMOSD, first-onset isolated double-negative myelitis, and unclassified inflammatory presentations. Prospective, multicenter longitudinal studies tracking serial clinical evaluations, standardized neuroimaging, and functional recovery will help clarify the natural history of these subgroups. Finally, researchers are investigating novel autoantibodies and biological profiles through cerebrospinal fluid proteomics and advanced cellular immunophenotyping. Validating stable candidate biomarkers in seronegative cohorts may eventually replace provisional serological labels with biologically defined, targeted disease mechanisms.

Subject of Research: Reclassification and diagnostic heterogeneity of antibody-stratified and double-seronegative acute inflammatory myelopathies

Article Title: Reclassification of acute inflammatory myelopathies in the antibody era: from AQP4-IgG and MOG-IgG to double-seronegative disease

Article References: Fan, P., Tan, Y., Ren, Z., & Zhang, Q. (2026). Reclassification of acute inflammatory myelopathies in the antibody era: from AQP4-IgG and MOG-IgG to double-seronegative disease. Frontiers in Neurology, 17, Article 1994489. https://doi.org/10.3389/fneur.2026.1994489

Image Credits: AI Generated

DOI: 10.3389/fneur.2026.1994489

Keywords: acute inflammatory myelopathy, double-seronegative disease, AQP4-IgG, MOG-IgG, neuromyelitis optica spectrum disorder, MOGAD, acute transverse myelitis, multiple sclerosis, cell-based assays, longitudinally extensive transverse myelitis, oligoclonal bands, GFAP astrocytopathy, neurosarcoidosis, spinal cord infarction, spinal dural arteriovenous fistula, cerebrospinal fluid pleocytosis, longitudinal reclassification