Dr Jasmin Patel
It is sometimes difficult to distinguish multiple sclerosis (MS) from other related neuroinflammatory diseases such as neuromyelitis optica spectrum disorder (NMOSD) and MOG-antibody associated disease (MOGAD) based on brain MRI. This is particularly an issue in clinical practice when there are multiple brain lesions present. Although MS lesions tend to occur in characteristic locations in the brain, which were described in “MS Lesion Checklist” (Kister, Practical Neurology, 2018) , it is unknown whether lesions in the locations from the Checklist can help distinguish MS from other neuroinflammatory “MS mimickers.” To answer this question, we selected a “development sample” of 82 patients with clinically definite MS, NMOSD, and MOGAD from NYU MS Care Center in New York who had at least 3 lesions on their brain MRI. Then three neuroradiologists blindly reviewed the scans and scored for the presence of lesions in the different locations from the Checklist. We then compared the frequencies of brain lesions in these locations in MS, NMOSD and MOGAD. We found that 3 locations were differentially affected in MS compared to NMOSD and MOGAD (“non-MS”) – periventricular (in ”Dawson’s finger” configuration), anterior temporal horn, and cerebellar hemisphere. Based on this, we developed a statistical prediction model using a multivariate logistic regression method that included these 3 lesion locations. We then tested our model using an independent “validation sample” of 97 different patients with MS, NMOSD, and MOGAD. In the validation sample as well, MS patients were more likely than non-MS patients to have lesions in all 3 locations identified in the development sample. Our model was 76% accurate in differentiating MS from “non-MS.” A simplified version of our model, which used combined validation and development samples, showed that the presence of lesions in at least 2/3 locations had 74.3% sensitivity and 79.1% specificity for diagnosing MS.
The major advantage of our approach is that it is readily applicable in the “real-world” setting – it only requires the clinician or radiologist to look for the presence of lesions in 3 locations on conventional brain MRI sequences, without the need for any advanced post-processing techniques, specialized sequences, or significant additional time. It should be noted that our algorithm is not meant to replace the validated criteria for MS or NMOSD, which are adequate for confident diagnosis in most cases, but can serve as an adjunct to these criteria in clinical scenarios where patients may meet McDonald Criteria and Barkhof criteria for dissemination in space in MS, and yet the diagnosis of MS is in question because of features that are atypical for MS.
We hope that our work will stimulate future studies to develop practical approaches for increasing diagnostic accuracy.
J. Patel, A. Pires, A. Derman, G. Fatterpekar, R.E. Charlson, C. Oh, I. Kister. Development and validation of a simple and practical method for differentiating MS from other neuroinflammatory disorders based on lesion distribution on brain MRI. Journal of Clinical Neuroscience, Volume 101, 2022, Pages 32-36, ISSN 0967-5868, https://doi.org/10.1016/j.jocn.2022.04.035.
There is an unmet need to develop practical methods for differentiating multiple sclerosis (MS) from other neuroinflammatory disorders using standard brain MRI. To develop a practical approach for differentiating MS from neuromyelitis optica spectrum disorder (NMOSD) and MOG antibody-associated disorder (MOGAD) with brain MRI, we first identified lesion locations in the brain that are suggestive of MS-associated demyelination (“MS Lesion Checklist”) and compared frequencies of brain lesions in the “MS Lesion Checklist” locations in a development sample of patients (n = 82) with clinically definite MS, NMOSD, and MOGAD. Patients with MS were more likely than patients with non-MS to have lesions in 3 locations only: anterior temporal horn (p < 0.0001), periventricular (“Dawson’s finger”) (p < 0.0001), and cerebellar hemisphere (p = 0.02). These three lesion locations were used as predictor variables in a multivariable regression model for discriminating MS from non-MS. The model had area under the curve (AUC) of 0.853 (95% confidence interval: 0.76-0.945), sensitivity of 87.1%, and specificity of 72.5%. We then used an independent validation sample with equal representation of MS and NMOSD/MOGAD cases (n = 97) to validate our prediction model. In the validation sample, the model was 76.3% accurate in discriminating MS from non-MS. Our simple method for predicting MS versus NMOSD/MOGAD only requires a neuroradiologist or clinician to ascertain the presence of lesions in three locations on conventional MRI sequences. It can therefore be readily applied in the real-world setting for training and clinical practice.
Disclaimer: The are the views of the author and no-one else