THU0652 Assessment of Biosimilars Using Real World Data: the Complexity of Choosing a Comparator and Understanding Uptake

Background The introduction of biosimilars has been linked with concerns regarding how to best monitor their similarity vs. the originator product using real world data. Methods Data from the Swedish Rheumatology Quality register (SRQ) was used to identify all patients with rheumatoid arthritis (RA), ankylosing spondylitis (AS), psoriatric arthritis (PsA), and other spondyloarthropathy (SpA), who started a treatment with infliximab (originator Remicade or biosimilars Remsima or Inflectra) between 1st Mar 2015 and 30th Sept 2016 or with etanercept (originator Enbrel or biosimilar Benepali) between 1st April 2016 and 30th Sept 2016. Results During the study period, a total of 1833 patients started an infliximab treatment and 1793 started etanercept. These patients were either bDMARD-naïve (patients without a history of any biological treatment), non-medical switchers (patients who switched from the originator product), or patients who had a history of a previous (but not the same) bDMARD (Table 1). These three groups were not evenly distributed across originators or biosimilars, and had different baseline demographic and disease characteristics. The uptake in terms of treatment starts was faster for Benepali (it covered more than 90% of this part of the etanercept market after only 3 months) as compared to Remsima and Inflectra (together they accounted for 88% of this section of the infliximab market after 10 months). The uptake of biosimilars in terms of proportion of all patient on treatment was, at the end of September 2016, 27% of all infliximab (after 18 months of marketing) and 16% of etanercept (after 5 months since marketing). In contrast to the bio-naïve group and those with a history of a previous (but not the same biologic), there was no readily available comparator group for the non-medical switcher group. To this end, we assessed three tentative definitions for a comparator; i) a historical comparison, i.e., same patients 18 months before the switch, ii) an individually matched sample of those patients still on originator treatment at the time of the switch, and iii) the total cohort of those who had not switched.Table 1. Number of patients starting infliximab (from 1 March 2015 to 30 September 2016) or etanercept (from 1 April 2016 to 30 September 2016) originator or biosimilar, by line of treatment Infliximab Etanercept Originator Biosimilars Originator Biosimilar Total 570 1263 186 1607 Bio- naïve 379 (66%) 508 (40%) 89 (48%) 581 (36%) Non-medical switchers – 524 (42%) – 710 (44%) Switchers from other biologics 191 (34%) 231 (18%) 97 (52%) 316 (20%)Figure 1. Number of patients starting, and on, respectively, infliximab or etanercept by month. Green: total, red: biosimilar, blue: originator. Conclusions “Uptake” of biosimilars can be expressed both as proportion of all new starts and as proportion of ongoing treatments. Assessments of uptake, and any comparison between biosimilars and their originators, need to be based on line of therapy in order to avoid mixing up effects of channeling with true differences between originator and similar. For the same reason, any originator comparator for non-medical switchers needs to be reflective of those patients who stood the same chances of switching, but did not switch. Disclosure of Interest D. Di Giuseppe: None declared, T. Frisell: None declared, S. Ernestam: None declared, H. Forsblad-d'Elia: None declared, E. Lindqvist: None declared, U. Lindström: None declared, C. Sjöwall: None declared, J. Askling Grant/research support from: Abbvie, BMS, Pfizer, MSD, Roche, Samsung, Lilly