Relapse Rates of Patients with Multiple Sclerosis Newly Initiating Subcutaneous Interferon ß-1a vs Oral Disease-Modifying Drugs in the Real World

Background: Administrative claims data provide information on outcomes in actual clinical care settings in a broader patient population. 

Objectives: To evaluate relapse rates of patients with multiple sclerosis (MS) newly initiating subcutaneous interferon β-1a (scIFNβ1a) vs oral disease-modifying drugs (DMDs; ie, teriflunomide, fingolimod, dimethyl fumarate) using real-world data.

Methods: Patients from the IMS LifeLink PharMetrics Plus™ Database between 1/1/2012 and 6/30/2013 met the inclusion criteria: MS diagnosis (ICD-9-CM:340.xx); initiation of scIFNβ1a, teriflunomide, fingolimod, or dimethyl fumarate (1st claim=index date); continuous eligibility 12 months pre- and post-index; no DMD 12 months pre-index (treatment-naïve); and age 18–63 years. Relapse was assessed 12 months following DMD initiation and was defined as: MS-related hospitalization, MS-related emergency room (ER) visit, or MS-related outpatient visit with corticosteroid prescription ±7 days. Analyses included pairwise chi-square tests and logistic regression controlling for age, sex, region, and clinically meaningful disease severity measures (ie, 90-day pre-index indicators for relapse, neurologist visits, and MRI). 

Results: 1665 patients (686 scIFNβ1a, 118 teriflunomide, 455 fingolimod, and 406 dimethyl fumarate) met inclusion criteria (mean age=44.4 years; 75.5% female). Unadjusted analyses showed that MS-related hospitalizations and ER visits did not differ among DMDs; however the proportion of patients with an MS-related outpatient relapse was lower in patients initiating scIFNβ1a (19.7%) vs teriflunomide (32.2%; p=0.003) and dimethyl fumarate (26.8%; p=0.006). Proportion of patients with ≥1 MS relapse of any type was lower with scIFNβ1a vs oral DMDs (21.7% and 26.1%, respectively, p = 0.039). Logistic regression controlling for demographic and 90-day pre-index indicators showed that initiation of teriflunomide or dimethyl fumarate were associated with higher likelihood of relapse (odds ratio [OR]=2.1; p=0.001 and OR=1.5; p=0.005, respectively) vs scIFNβ1a. A neurologist visit (p=0.034) and MS relapse (p<0.0001) 90 days before treatment initiation were predictive of relapse.

Conclusions: In this real-world population, after controlling for demographics and pre-index clinically meaningful indicators of disease severity, patients initiating scIFNβ1a had a lower likelihood of experiencing surrogates for relapse in the first year than patients initiating teriflunomide or dimethyl fumarate.