Treating Relapsing Multiple Sclerosis with Subcutaneous versus Intramuscular Interferon-Beta-1a

Abstract

The EVIDENCE trial concluded that administering high-dose/high-frequency subcutaneous (SC) interferon-beta-1a (IFNb1a) was more effective in preventing relapses among patients with relapsing multiple sclerosis (MS) than low-dose weekly intramuscular (IM) IFNb1a after 64 weeks. This analysis utilized discrete-event simulation (DES) to model the potential longer-term clinical and economic implications of this trial. A DES predicting the course of relapsing MS and incorporating the effect of IFNb1a therapy was developed. The model began by randomly reading in actual patient data from the trial to create 1000 patients. Each simulated patient was replicated - one was assigned to receive SC IFNb1a three times a week and the other to receive IM IFNb1a once a week. During the simulation, patients may (i) experience relapses, with associated short- and long-term impacts on costs and disability; (ii) develop new T2 lesions detected by a magnetic resonance imaging scan; (iii) discontinue treatment because of adverse events or lack of response; (iv) advance to secondary progressive MS; or (v) die. Model inputs were mainly obtained from the EVIDENCE trial, but were taken from published literature if they could not be obtained from the trial. Direct medical costs ($US, year 2006 values) to the US payers were primarily obtained by updating a published cost analysis. Costs and benefits were discounted at 3% per annum. Extensive sensitivity analyses were conducted to test the robustness of the model results. Based on 100 replications of 1000 patient pairs over 4 years, SC IFNb1a was predicted to enable more patients to avoid relapse (216 vs 147). Total mean costs per patient (discounted) were $US79 890 with SC IFNb1a versus $US74 485 with IM administration, a net increase of $US5405 per patient. However, SC IFNb1a was estimated to prevent 0.50 relapses and save 23 relapse-free days per patient, yielding incremental cost-effectiveness ratios of $US10 755 per relapse prevented and $US232 per relapse-free day gained. Sensitivity analyses revealed that the result was most sensitive to the treatment efficacy, model time horizon and cost of IFNb1a treatment. Based on the results observed in the EVIDENCE trial, the model predicted that SC IFNb1a would yield greater health benefits over 4 years than IM IFNb1a, at a cost that would seem to be a reasonable trade-off.