Real-world Digital Biomarkers Associated With Their Clinical Counterparts in the MS-DETECT Study
Patrick Vermersch1, Massimo Filippi2, Celia Oreja-Guevara3, Jiwon Oh4, Tobias Sejbæk5, Jennifer Graves6, Lissandra Klaeylé7, Séverine Bieuvelet7, Loïc Carment7, Pierre DROUIN7, Saad Zinaï7, Pascal Rufi8, Bruno Padrazzi8, Tjalf Ziemssen9
1Univ. Lille, UMR Inserm U1172, CHU Lille, Lille, France, 2Neurology Unit, IRCCS San Raffaele Scientific Institute, Milan, Italy, 3Hospital Clinico San Carlos, IdISSC, Madrid, Spain, 4Division of Neurology, St. Michael’s Hospital, University of Toronto, Toronto, Canada, 5Department of Neurology, Esbjerg Hospital, University Hospital of Southern Denmark, Esbjerg, Denmark, 6Department of Neurosciences, University of California San Diego, San Diego, CA, USA, 7Ad Scientiam, Paris, France, 8Sanofi, Paris, France, 9MS Center, Center of Clinical Neuroscience, Neurological Clinic, University Clinic Carl Gustav Carus, TU Dresden, Dresden, Germany
Objective:
This interim analysis describes baseline comparisons between MSCopilot® digital biomarkers, measured in-clinic or at-home, and individual MSCF-4 scores.
Background:
Standard clinical assessments lack the sensitivity to capture subtle disease progression in people living with multiple sclerosis (PwMS). MSCopilot®, a Software as a Medical Device, remotely assesses key functions such as cognition, dexterity, vision and walking. Previous French studies showed that the MSCopilot® scores correlate with Extended Disability Status Scale (EDSS) and the revised MS Functional Composite (MSFC-4). The international MS-DETECT study (NCT05816122) evaluates MSCopilot®’s ability to detect disability worsening in PwMS during 18-24 months.
Design/Methods:
Patients’ clinical data, including their MSFC-4 assessments, were collected during the inclusion visit. Patients performed MSCopilot® tests in-clinic and later at home. The association between the in-clinic and remote digital biomarkers scores was assessed using intraclass correlation coefficients and association with their MSFC-4 counterparts was assessed with Pearson correlation coefficients.
Results:
244 (68.4% female, mean age 49.15 ± 8.74, mean disease duration 17.57 ± 8.04 and mean EDSS: 3.8 ± 1.3) patients were included in this analysis. We collected 925 in-clinic and 829 remote MSCopilot® evaluations (243/218 cognition, 231/211 dexterity, 209/182 vision and 242/218 walking). Digital biomarkers measurements were consistent with good (0.64) to excellent (0.77) intraclass coefficients. Moderate to strong correlations were found between individual MSFC-4 items and digital tests: cognition (0.71/0.62), dexterity (0.45/0.47) and walking (0.54/0.66) tests. Composite MSFC-4 scores also correlated with MSCopilot® 's measured in-clinic (0.71, p<0.001) and at home (0.71, p<0.001).
Conclusions:
MSCopilot® digital biomarkers met excellent quality standards with consistent measures correlating with clinical counterparts. These findings emphasize their potential use in detection of disease worsening leading to improved clinical monitoring of MS.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.