Background/Purpose: According to national guidelines issued in May 2015, a non-medical switch from originator infliximab (IFX) (Remicade) to biosimilar Remsima was conducted in all Danish patients with inflammatory rheumatic diseases treated in routine care. We aimed to investigate the clinical outcomes in Remicade-treated patients (pts) with rheumatoid arthritis (RA), psoriatic arthritis (PsA) and axial spondyloarthrits (AxSpA) who were switched to Remsima and monitored prospectively in the Danish nationwide quality registry, DANBIO.

Methods: Patients who switched from Remicade to Remisma before Feb 1^st 2016 and who had available data on treatment outcomes in DANBIO were included.  In each patient, disease activity at 3 mths before switch (pre-switch) (defined as between minus 180 and minus 30 days before switch), at the time of switch (-30 days before to 6 days after), after 3 mths (70 to 120 days after switch) (post-switch) and changes over time (Dpre-switch and Dpost-switch) were calculated. Disease flare was defined as DDAS28≥0.6, DDAS28≥1.2 (in RA and PsA) or DASDAS≥1.3 (AxSpA). Reasons for withdrawal (adverse events (AE), lack of effect (LOE) or other) were registered. Multivariable Cox regression analyses stratified according to diagnosis (RA, PsA or AxSpA) were used to identify baseline characteristics (gender, age,  use of methotrexate (yes-no),  Remsima dosing interval, Remsima dose (mg/kg),  patient’s global score (Visual analogue scale from 0-100mm, VAS)) associated with Remsima treatment adherence.  Numbers are medians (interquartile ranges) unless otherwise stated.

Results: 768 of 792 switching pts (97%) had available data regarding disease activity after switching. 52% were women, age (56 (46-66)yrs) (Table). Prior Remicade treatment duration was 6.6 (3.9-9.4) yrs, and in 75% of patients it was the first biological treatment. Median follow-up time after switching was 336 (297-357) days. Disease activity and disease flare rates remained largely unchanged 3 months prior to versus after the switch (Table 1). At the latest visit after switch, disease activity was for RA: DAS28 2.2 (1.7-3.0) and for axSpA: ASDAS 2.0 (1.2-2.9). Proportion of patients with disease flare pre-/post switch was 25%/21% (DDAS28≥0.6, RA+PsA)(p=0.4), 10%/10% (DDAS28≥1.2,RA+PsA) (p=0.8) and 2%/2% (AxSpA)(p=1.0) (related samples McNemar test). Overall, 117 patients (15%) stopped Remsima treatment between switch and end of follow-up (AE 34, LOE 51, remission 1, death 2, cancer 1, pregnancy 1, other 27). Prior Remicade treatment duration in those who withdrew was 6.0 (2.6-8.9) years and disease activity in those who stopped due to LOE was for RA: DAS28 4.3 (3.1-4.9) and for AxSpA: ASDAS 3 (2.6-3.8). Treatment adherences were similar in RA, PsA and AxSpA (Figure). In multivariable Cox regression analysis, lower baseline VAS patient’s global and use of methotrexate was associated with better treatment adherence in RA (Table 2) whereas in AxSpA it was lower baseline VAS global and lower IFX doses (Table 2). No significant baseline predictors were found in PsA.

Conclusion: In 768 patients with inflammatory rheumatic diseases treated with Remicade for >4 years, disease activity was largely unaffected in the majority of patients after non-medical switch to biosimilar Remsima, and the fluctuations 3 months after the switch were comparable to the fluctuations observed in the 3 months prior to the switch.  However, several patients (15%) stopped treatment. This warrants further investigation before such a non-medical switch can be recommended.