OBJECTIVES : Multiple sclerosis (MS) is classified into three subtypes; relapsing-remitting (RRMS), secondary progressive (SPMS; a progression from RRMS), and primary progressive (PPMS; no prior relapsing-remitting history). Many RRMS treatments have undergone HTA and robust methodologies for cost-effectiveness modeling have been developed. We assess use of RWE in economic modelling for PPMS.

METHODS : As of January 2018, one HTA has been published in the US for PPMS by the Institute for Clinical and Economic Review (ICER). This study reviewed RWE used in the cost-effectiveness models developed for ICER assessment of disease modifying therapies in MS.

RESULTS : For RRMS, ICER adapted an established model to include 10 health states for RRMS based on the level of disability, quantified using the Expanded Disability Status Scale (EDSS) (EDSS 0‒9), transition to SPMS, 9 health states for SPMS (EDSS 1‒9), and death. The PPMS model further adapted this structure, including 9 health states (EDSS 1‒9) and death. In contrast to the RRMS model, patients with PPMS were assumed not to experience relapse. RWE included baseline EDSS, transition probabilities, utilities, and costs by EDSS state. Notably, transition probabilities for RRMS, RRMS to SPMS, and SPMS were obtained from RCTs, supplemented with RWE from a Canadian cohort. In the absence of RWE for PPMS, ICER assumed that transition probabilities would be equal to SPMS. Utilities were obtained from RCTs (RRMS) and UK RWE (SPMS), while for PPMS ICER assumed that utilities would be equal to SPMS.

CONCLUSIONS : The absence of RWE for PPMS led to assumptions of equivalence to SPMS. Whilst both are progressive forms of MS, factors including prior treatment (more likely for SPMS) and age at MS diagnosis (typically older for PPMS) may impact the validity of these assumptions. HTA agencies such as NICE may require PPMS-specific RWE.