“Interestingly, when you compare one patient to another, the most important predictors are different. As clinicians, we often say that for this patient, what really matters is the duration of the medication. Or, in this patient’s case, it might be age. We found that in developing this model. This helps reiterate that different factors are influencing each individual.”
Recently, researchers around the world have cleveland clinic We observed that current guidelines for multiple sclerosis (MS) do not seem to provide an evidence-based approach to discontinuing disease-modifying therapies (DMTs). To address this gap, they recently developed a machine learning-based tool to predict the individualized risk of MS disease activity relapse after DMT discontinuation. The model was derived from a retrospective cohort of adults with MS who attended the clinic at least twice from January 2015 to July 2023 and was externally validated using data from the DISCOMS trial (NCT03073603). The primary endpoint was MS inflammatory activity, including relapse, new T2 lesions, or gadolinium-enhancing lesions, assessed after 2 years.
The analysis published in 2025 European Committee for Treatment and Research in Multiple Sclerosis (ECTRIMS) Conferenceheld in Barcelona, Spain, September 24-26, and involved a total of 1104 patients from the development cohort (withdrawn, n = 184) and 259 DISCOMS participants (withdrawal, n = 131). Presentation by first author Marisa McGinley, DOthe model performed similarly in both cohorts, predicting an average risk of recurrent disease activity for discontinuers of 12.6%. Researchers reported important predictors including months since inflammatory activity, duration of current DMT use, age, months since last relapse or MRI activity, and DMT efficacy.
After the presentation, McGinley, a staff neurologist at the MS Mellen Center at the Cleveland Clinic, said: neurology live® She discusses how her team used an individualized risk prediction model for MS patients considering DMT discontinuation. He noted that the model provides patient-specific predictions and highlights how various factors, such as age, disease duration, and treatment history, affect an individual’s risk. Additionally, McGinley emphasized that the tool supports shared decision-making, rather than dictating treatment choices, and provides clinicians with concrete data to guide treatment discussions with patients.
Click here for more information about ECTRIMS 2025.
References
1. McGinley M, Felix C, Lapin B, Corboy J, Ontaneda D, et al. A personalized decision support tool to predict the risk of relapse of disease activity after discontinuing disease-modifying therapy in multiple sclerosis. Presented at the ECTRIMS conference. September 24-26, 2025. Barcelona, Spain. Abstract O119.
