INFORMING MEDICATION DISCONTINUATION DECISIONS AMONG ADULTS WITH MULTIPLE SCLEROSIS

Purpose: To estimate the effects of discontinuing medication among adults with clinically stable relapse-remitting multiple sclerosis (RRMS). Medications for RRMS that target the immune system (disease-modifying treatments [DMTs]) are burdensome to patients due to adverse effects, frequent injections, and high costs. Understanding if and when DMTs can be safely discontinued is an important question for many people with RRMS and their neurologists.

Method: We developed a Markov model that incorporates the concept of disease heterogeneity (i.e., some people face higher risks than others). We modeled heterogeneity in the risks of relapse and progression with three subgroups in a simulated cohort of newly-diagnosed RRMS patients.  Subgroup-specific parameters were calibrated to outcomes from a large, population-based published longitudinal data from the British Columbia MS (BCMS) database. For example, the percent of the study cohort that experienced a five-year relapse-free interval provided a useful calibration target for heterogeneity in relapse rates (i.e., if all simulated patients had the same relapse risk, the percent who experienced a five-year relapse-free interval predicted by the model would not have matched the study). Average progression and relapse rates (across all subgroups) and other model parameters were derived from the literature. We used our calibrated model to identify a cohort of older patients who were clinically stable in mild disease after long-term treatment, and characterized the underlying heterogeneity of this cohort. We then compared quality-adjusted life years (QALYs) for continuing vs. discontinuing DMTs in a range of clinical scenarios (based on age and individual preferences). 

Result: Choosing to discontinue DMTs for people with RRMS who were experiencing mild disease at age 55 or 60 (no relapses or progression for at least five years) reduced QALYs by about three quality-adjusted days. This converts into a treatment disutility threshold of 0.995 utility for being on treatment. This small benefit might not be worth the burden of taking medication for some patients.

Conclusion: Among older adults with clinically stable RRMS on long-term DMTs, we found that the benefits of continuing treatment were small. These results suggest that better understanding if and when to stop medication can improve the quality of life for people with RRMS. Our results may help people with RRMS make treatment decisions in a manner that is consistent with their preferences and values.