Pos1549 better drug survival on monoclonal antibodies compared to etanercept in first line anti-tnf therapy in patients with psoriatic arthritis may be due to differential therapeutic response in skin involvement: a propensity score-matched analysis from the nationwide czech attra registry

Background Etanercept is an agent of specific molecular structure and its efficacy in extraarticular disease differs from monoclonal antibodies. This could affect drug survival in some rheumatic diseases with a larger presence of extraarticular involvement. Objectives To compare the drug survival of etanercept to monoclonal anti-TNF antibodies in psoriatic arthritis and describe the response of skin disease to treatment in the two groups. Methods Patients with PsA starting first-line biological treatment from 01/01/2012 to 30/06/2020 were enrolled from the Czech ATTRA registry and split into two groups as either being treated with etanercept (ETA) or any other anti-TNF. Propensity scores were calculated using covariates based on statistically significant differences in the baseline characteristics and clinical relevance (Table 1). PS was then used to match 81 etanercept to 160 other TNFi patients. We performed a Kaplan-Meier survival analysis and checked for statistical significance using the log-rank test. We calculated survival rates at set time points, median survival time in each group and the hazard ratio of etanercept for drug discontinuation. The Physician global assessment of psoriasis variable (measured on a numerical rating scale 0-5) was sorted in categories of mild (0-1), intermediate (2-3) and severe (4-5) and respective percentages were calculated in the propensity-matched groups at 3, 6, 12 and 24 months. Pearson’s chi-squared test or Fisher’s exact test were used to check the differences for statistical significance. Results We found significantly worse drug survival on ETA compared to other TNFi at each time point (Figure 1). The median survival time on ETA was 35.8 (95% CI 25.1 – 46.6) months compared to 65.7 (95% CI 51.9 – 79.6) months on other TNFi. The HR of ETA for treatment termination was 1.61 (95% CI 1.11 - 2.34), p=0.011. At baseline, there were no differences in skin disease severity. The percentages of mild, intermediate and severe skin disease were 22.2%, 59.3% and 18.5% on etanercept and 25.0%, 55.6% and 19.4% on other TNFi, p=0.853. At 3 months, skin disease was significantly more severe in ETA with 50.0%, 44.8% and 5.2% compared to 76.2%, 23.8% and 0% on other TNFi, p<0.001. At 12 months, the respective percentages were 66.6%, 31.9% and 2.1% on ETA and 84.7%, 15.3% and 0% on other TNFi, p=0.015; at 24 months 62.9%, 37.1% and 0% compared to 84.9%, 15.1% and 0%, p=0.010. Conclusion Etanercept had worse drug retention than monoclonal antibodies in PsA. Patients treated with ETA had more severe skin disease during follow-up compared to those treated with other TNFi, which could lead to earlier switching to another drug. Figure 1. Drug survival of etanercept vs. other TNF inhibitors in patients with PsA. Table 1. Selected baseline characteristics after propensity score matching. Etanercept (n=81) Other TNFi (n=160) p Females 42 (51.9%) 78 (48.8%) 0.649 Age at diagnosis 40.0 (34.0–50.0) 42.0 (34.0–49.0) 0.847 Age at 1st line treatment 51.0 (42.0–59.0) 52.0 (44.0–59.0) 0.835 Disease duration (yrs) 7.1 (2.3–14.4) 7.8 (2.9–15.2) 0.587 DAPSA 36.3 (28.1–43.6) 36.2 (26.7–45.8) 0.946 CRP (mg/dl) 12.0 (4.8–25.0) 15.4 (6.3–28.0) 0.227 ESR (mm/h) 26.0 (16.0–37.0) 30.0 (15.0–40.0) 0.728 Tender joint count (68) 12.0 (9.0–19.0) 12.0 (7.0–19.0) 0.602 Swollen joint count (66) 8.0 (5.0–11.0) 8.0 (4.0–12.0) 0.759 Patient global assess. (0-100) 68.0 (50.0–80.0) 70.0 (50.0–80.0) 0.832 Physician global assess. (0-100) 60.0 (45.0–70.0) 60.0 (38.5–72.5) 0.942 Yr of administration 2012-2013 7 (8.6%) 15 (9.4%) 0.973 2014-2015 16 (19.8%) 35 (21.9%) 2016-2017 21 (25.9%) 41 (25.6%) 2018-2020 37 (45.7%) 69 (43.1%) Concomitant csDMARD 66 (81.5%) 126 (78.8%) 0.619 Concomitant MTX 53 (65.4%) 97 (60.6%) 0.467 Concomitant glucocorticoids 25 (30.9%) 54 (33.8%) 0.652 Median (IQR) in continuous variables, n (percentage) in categorical variables. Acknowledgements This work was supported by the project (Ministry of Health, Czech Republic) for conceptual development of research organization 00023728 (Institute of Rheumatology). Disclosure of Interests None Declared.