The management of GI symptoms associated with the initiation of DMF has evolved over time with real world experience. The combined effect of GI management strategies has not been assessed in randomized studies and their overall impact on the incidence of GI symptoms and therapy discontinuation due to GI events is currently unclear.
To compare discontinuation rate due to gastrointestinal (GI) symptoms in the first year after initiating dimethyl fumarate (DMF) between an early (post launch) and recent DMF cohort of multiple sclerosis (MS) patients at two MS centers.
Data were collected through chart reviews at Rocky Mountain MS Center and Providence MS Center. The cutoff for Early versus Recent cohort was April 1, 2014. Chi-square, non-parametric, and t-tests were used to determine differences between the two cohorts. Differences in discontinuation due to a GI event for the two cohorts were compared with a Cox proportional hazards model adjusted for baseline characteristics.
Medical records of 700 patients who initiated DMF between March 2013 and December 2017 were reviewed- 302 were Early and 398 Recent. At baseline, Early patients were older (50.30 vs 48.50, p=0.049), had longer disease duration [11.09 (IQR: 7.88, 16.77) vs 6.72 (4.35, 13.72), p<0.001], more history of GI disease (27.8% vs. 22.1%, p=0.099) but lower percent of seasonal allergy (17.9 vs. 24.4, p=0.048). Discontinuation for any reason was higher in Recent patients (37.4% vs. 22.2%, p<0.001). Discontinuation due to GI symptom were 8.6% in the Early patients and 12.1% in the Recent patients (p=0.178). Females [HR: 2.27 (1.12-4.60)] and patients with a history of GI condition [HR: 2.14, (1.32-3.47)] were more likely to discontinue. No significant associations were found between discontinuation and any GI events or previous exposure to natalizumab or other medications. None of the baseline or patient characteristics were found to be risk factor of GI events except study site.
Overall, the discontinuation rate due to GI symptoms was no different between the Early and Recent patient cohorts. However, the odds of having a GI symptom at the Rocky Mountain site were higher although discontinuation for this site was lower suggesting there may be differences in symptom reporting, data abstraction, and GI mitigation strategies between the two centers. Further research with prospective study design and standard documentation and data abstraction tool are needed.
Support: Biogen