MRI Criteria Show Diagnostic Accuracy for Distinguishing MS From NMOSD and MOGAD

Lesion distribution criteria by magnetic resonance imaging (MRI) accurately differentiates multiple sclerosis (MS) from neuromyelitis optica spectrum disorders (NMOSD) and myelin oligodendrocyte glycoprotein antibody-associated disease (MOGAD), according to findings of a review and meta-analysis published in Journal of Neurology, Neurosurgery, and Psychiatry.

Researchers evaluated the effectiveness of MRI lesion criteria to distinguish between MS, NMOSD, and MOGAD across various clinical settings (PROSPERO Number: CRD42023472178). They conducted a search of online databases for studies including patients with MS, NMOSD, and MOGAD. The analysis focused on lesion distribution using brain MRI data. The primary outcomes were measures of diagnostic accuracy.

A total of 11 studies published between 2013 and 2022, involving 2008 participants, were included in the analysis. Of the participants, 1037 had MS, 842 had NMOSD, and 129 had MOGAD. Lesion distribution criteria were evaluated, with 9 studies applying the 2013 Matthews criteria to differentiate between MS, NMOSD, and MOGAD. Other studies used the 2019 Cacciaguerra criteria or the MS lesion checklist. Lesion distribution was assessed in different populations and clinical settings. Eight studies reported “satisfying agreement” in inter-rater reliability, with raters who were neurologists and neuroradiologists.

Lesion distribution criteria present high diagnostic accuracy for discrimination between MS, NMOSD and MOGAD, proving their potential in complex cases that often lead to misdiagnosis and treatment delay.

For distinguishing MS from NMOSD, 9 studies using the Matthews brain MRI criteria showed a pooled sensitivity of 92% and specificity of 85%. Among these, 4 studies that specifically validated the criteria reported pooled sensitivity and specificity of 93% and 88%, respectively.

In differentiating MS from MOGAD, 5 studies reported both pooled sensitivity and specificity at 86%. The Cacciaguerra brain-spinal cord criteria, used in 4 studies to distinguish MS from NMOSD, showed pooled sensitivity of 96% and specificity of 83%. In contrast, the MS lesion checklist demonstrated lower diagnostic accuracy for differentiating MS from NMOSD/MOGAD, with sensitivity of 74% and specificity of 79%.

Subgroup analyses for the Matthews criteria showed pooled sensitivity of 92% in White patients and 94% in non-White patients, with specificity of 83% vs 87%, respectively. The Matthews criteria had higher utility for MS confirmation in White patients, while in Asian patients, they were effective for both confirming and ruling out MS. When examining performance based on timing, the Matthews criteria demonstrated a pooled sensitivity of 92% and specificity of 84% at disease onset, and a sensitivity of 91% and specificity of 90% at follow-up.

Overall, the Matthews criteria provided the strongest moderate certainty evidence and the highest pooled diagnostic accuracy for differentiating MS from NMOSD (sensitivity, 93%; specificity 90%) and from MOGAD (sensitivity, 86%; specificity, 87%).

Study limitations include heterogeneity within and between the studies.

“Lesion distribution criteria present high diagnostic accuracy for discrimination between MS, NMOSD and MOGAD, proving their potential in complex cases that often lead to misdiagnosis and treatment delay,” the researchers noted.

Disclosures: Multiple authors declared affiliations with biotech, pharmaceutical, and/or device companies. Please see the original reference for a full list of authors’ disclosures.