Patterns and predictors of multiple sclerosis phenotype transition

Abstract

Currently, there are limited therapeutic options for patients with non-active secondary progressive multiple sclerosis. Therefore, real-world studies have investigated differences between patients with relapsing-remitting multiple sclerosis, non-active secondary progressive multiple sclerosis and active secondary progressive multiple sclerosis. Here, we explore patterns and predictors of transitioning between these phenotypes. We performed a cohort study using data from The Danish Multiple Sclerosis Registry. We included patients with a relapsing-remitting phenotype, registered changes to secondary progressive multiple sclerosis and subsequent transitions between relapsing and non-relapsing secondary progressive multiple sclerosis, which was defined by the presence of relapses in the previous 2 years. We analysed predictors of transitioning from relapsing-remitting multiple sclerosis to relapsing and non-relapsing secondary progressive multiple sclerosis, as well as between the secondary progressive states using a multi-state Markov model. We included 4413 patients with relapsing-remitting multiple sclerosis. Within a median follow-up of 16.2 years, 962 were diagnosed with secondary progressive multiple sclerosis by their treating physician. Of these, we classified 729 as non-relapsing and 233 as relapsing secondary progressive multiple sclerosis. The risk of transitioning from relapsing-remitting to non-relapsing secondary progressive multiple sclerosis included older age (hazard ratio per increase of 1 year in age: 1.044, 95% confidence interval: 1.035-1.053), male sex (hazard ratio for female: 0.735, 95% confidence interval: 0.619-0.874), fewer relapses (hazard ratio per each additional relapse: 0.863, 95% confidence interval: 0.823-0.906), higher expanded disability status scale (hazard ratio per each additional point: 1.522, 95% confidence interval: 1.458-1.590) and longer time on disease-modifying therapies (hazard ratio per increase of 1 year in treatment, high-efficacy disease-modifying therapy: 1.095, 95% confidence interval: 1.051-1.141; hazard ratio, moderate-efficacy disease-modifying therapy: 1.073, 95% confidence interval: 1.051-1.095). We did not find significant predictors associated with the transition from relapsing secondary progressive multiple sclerosis to non-relapsing secondary progressive multiple sclerosis, whereas older age (hazard ratio per increase of 1 year in age: 0.956, 95% confidence interval: 0.942-0.971) prevented the transition from non-relapsing secondary progressive multiple sclerosis to relapsing secondary progressive multiple sclerosis. Our study suggests that transitioning from relapsing-remitting multiple sclerosis to non-relapsing secondary progressive multiple sclerosis depends on well-known factors affecting diagnosing secondary progressive multiple sclerosis. Further transitions between non-relapsing and relapsing secondary progressive multiple sclerosis are only affected by age. These findings add to the knowledge of non-active secondary progressive multiple sclerosis, a patient group with unmet needs in terms of therapies.

Keywords: multiple sclerosis; real-world data; registry study; secondary progressive multiple sclerosis.

Graphical Abstract

Figure 1

Patients’ disposition flowchart.

Figure 2

Example of multiple sclerosis phenotype identification procedure. The example uses an RRMS patient later diagnosed with SPMS. The multiple sclerosis phenotype is RRMS from the day of multiple sclerosis diagnosis until the day of SPMS diagnosis. At this day (indicated as Day 1 in the ‘Day of SPMS phenotype assessment’ timeline), the patient is deemed rSPMS because a relapse (R) overlaps with the 2-year ‘retrospective time window Day 1’. The patient multiple sclerosis phenotype remains rSPMS until Day 630, when it is deemed nrSPMS, as the application of the retrospective 2-year time window does not overlap with a relapse. At Day 6547, the patient transition again to rSPMS after experiencing a relapse and stays in this category until the end of follow-up (Day 6721).

Figure 3

Aalen–Johansen cumulative incidence functions estimate stacked plots of multiple sclerosis phenotype transitions in the selected population. (A) Cumulative risk of receiving an SPMS diagnosis with a non-relapsing (violet) or relapsing (orange) course, staying event-free (black) or dying (grey) for the initial RRMS population (n = 4413) over a maximum follow-up time of 22.8 years. (B) Cumulative risk of ever transitioning to a rSPMS phase, dying or remaining relapse-free for the nrSPMS population at SPMS diagnosis (n = 962) over a maximum follow-up time of 15 years since SPMS diagnosis. (C) Cumulative risk of ever transitioning to an nrSPMS phase, dying or never entering a nrSPMS for the rSPMS population at SPMS diagnosis (n = 233) over a maximum follow-up time of 7.8 years since SPMS diagnosis. In all panels, the distances between two adjacent curves represent the cumulative probabilities, expressed as percentages, of the different events.