Related Experiment Video
Updated: May 27, 2026
Quantification of Protein Interaction Network Dynamics using Multiplexed Co-Immunoprecipitation
Published on: August 21, 2019
A clustering-based approach to characterize autonomy profiles among multiple sclerosis patients: an application of
David Pau1, Alexandre Civet2, Cécile Donzé3
1Science, Roche SAS, Paris, France. david.pau@roche.com.
Background And Objectives:
Many Multiple Sclerosis (MS) patients have symptoms that impact on their autonomy, defined as being able to perform the roles that are most important to oneself, with or without help. This research aimed to identify distinct phenotypic autonomy profiles among People with MS (PwMS) and evaluate their 1-year longitudinal trajectories using the newly developed 10-dimension Multiple Sclerosis Autonomy Scale (MSAS).
Methods:
Data were drawn from the prospective, non-interventional FOCAL-MS2 study following 199 PwMS in France over 12 months. To uncover patient profiles, we applied an unsupervised clustering method using clinical and social parameters at baseline. Cluster stability was validated via bootstrap resampling and the Jaccard similarity index. Within-patient 1-year changes in MSAS subscales were evaluated against a Minimum Clinically Important Difference (MCID) threshold.
Results:
The analysis yielded six highly stable patient clusters (Jaccard indices ranging from 0.90 to 0.98). Cluster 1 (n = 30) comprised middle-aged patients on sick leave who were severely impacted regarding activities with others. Cluster 2 (n = 42) was characterized by a younger population (40-49 years) requiring occasional walking assistance. Cluster 3 (n = 42) consisted of middle-aged patients who maintained high levels of personal activities. Cluster 4 (n = 30) represented the oldest (≥60 years) and most disabled group at baseline; paradoxically, this subgroup showed an improvement in autonomy across most dimensions over the year. Finally, Cluster 5 (n = 19, without activities) and Cluster 6 (n = 36, professionally active women) both represented younger patients (<40 years) exhibiting high functional autonomy at baseline.
Conclusion:
Our study identified six unique clusters that differ not only in their baseline symptom burden but also in their 1-year trajectory of autonomy. This underlines the necessity to address specific socio-professional and psychosocial dimensions of autonomy unique to each patient's profile.
