POS0940 FACTORS ASSOCIATED WITH LONG-TERM RETENTION OF TREATMENT WITH GOLIMUMAB IN A LARGE COHORT OF PATIENTS WITH RHEUMATIC DISEASES, WITH UP TO 8 YEARS OF FOLLOW-UP

Abstract

BackgroundThe long-term retention rate of a biological drug is a surrogate marker of its effectiveness and tolerability.ObjectivesWe assessed the probability of golimumab retention (persistence or drug survival) and the associated factors in a large cohort of patients with rheumatic diseases, with up to 8 years of follow-up.MethodsThis was an analysis of the BIOBADASER database (Spanish registry of biological drugs of the Spanish Society of Rheumatology and the Spanish Medicines Agency) on all adult patients who had initiated golimumab for treating rheumatoid arthritis (RA), psoriatic arthritis (PsA) or axial spondyloarthritis (SpA). The probability of golimumab retention was assessed with the Kaplan-Meier method, differences between groups with the log-rank test, and factors related to retention with a Cox-regression model. Patients were right-censored if they were still treated with golimumab at the last observation for data analysis.ResultsA total of 885 patients were included, of whom 59 had received golimumab in 2 separate cycles (with a grace period of 3 months), totaling 944 cycles of treatment (286 RA, 396 axial SpA and 262 PsA). At golimumab initiation, mean (SD) age was 52 (13) years, 54% were women and median duration of disease was 7.6 (2.8-14.4) years. Golimumab was prescribed as first, second and third/subsequent biological drug in 313 (33%), 303 (32%) and 328 (35%) treatments. Concomitant medications at golimumab initiation included methotrexate (MTX) (32%), steroids (29%), leflunomide (13%) and sulphasalazine (6%). The probability of retention of golimumab since treatment initiation was 71% (95% confidence interval [CI]: 68 – 74) at year 1, 60% (95% CI: 57-63) at year 2, 54% (95% CI: 51-58) at year 3, 48% (95% CI: 44-51) at year 4, 44% (95% CI: 40-48) at year 5, 41% (95% CI: 37-45) at year 6 and 38% (95% CI: 33-42) at year 7 and at year 8. In bivariate analysis, the retention rate was higher when golimumab was used as first biological agent (p log-rank <0.001), in patients with axial SpA or PsA compared to RA (p <0.001, Figure 1) in men (p<0.001) and in those not using steroids (p<0.001). As first biological drug the probability of retention was 82% at year 1 and 50% at year 7. As second, it was 70% at year 1 and 35% at year 8, and at third/further, 62% (year 1) and 28% (year 8). Cox-regression analysis (Table 1) showed that golimumab retention was positively associated to use as first vs second or third biological, to axial SpA or PsA compared to RA and to use of methotrexate, and inversely to corticosteroid use and to disease activity over the median at golimumab initiation.Table 1.Cox-regression analysis. Hazard Ratio for discontinuation of golimumabHazard Ratio95% Confidence intervalpAge at golimumab initiation1.011.00-1.020.063Gender (women vs men)1.230.98-1.550.079Axial SpA vs RA0.590.44-0.80<0.001PsA vs RA0.670.51-0.890.005Second vs first biological drug1.521.17-1.970.002Third or further vs first biological drug1.791.38-2.32<0.001Corticosteroids1.461.16-1.850.001Methotrexate0.790.63-0.990.041Disease activity over the median*1.291.05-1.590.015*DAS28 > 4.3 (RA, PsA) or BASDAI > 5.6 (axial SpA) at golimumab initiationFigure 1.ConclusionThis study provides new information on long-term golimumab effectiveness for the treatment or rheumatic diseases, with retention rates of 71% at year 1, 44% at year 5 and 38% at year 8. The probability of golimumab retention was higher as first biological drug, in patients with PsA or axial SpA and in those treated with methotrexate, and lower in those treated with steroids or with higher disease activity at golimumab initiation.AcknowledgementsBIOBADASER is funded by the Spanish Society of Rheumatology, the Spanish Agency of Medicines and by different pharmaceutical companies. The present study was funded by MSD, Spain.have no acknowledgements to declare.Disclosure of InterestsManuel Pombo-Suarez: None declared, Daniel Seoane-Mato: None declared, Luis Cea-Calvo Employee of: Medical Affairs, MSD Spain, Federico Diaz-Gonzalez: None declared, Fernando Sánchez-Alonso: None declared, Marta Sánchez-Jareño Employee of: Medical Affairs, MSD Spain, Francisco Javier Manero Ruiz: None declared, Lucía Ruiz: None declared, Vega Jovani: None declared, Isabel Castrejon: None declared