Background As in rheumatoid arthritis some patients with psoriatic arthritis (PsA) can be refractory to anti-TNF-α therapy, may lose responsiveness, or develop drug intolerance over time and a switch to another TNFα antagonist can often restore therapeutic response.

Objectives The study aim was to determine treatment persistence and to identify causes of discontinuation in PsA patients in clinical practice.

Methods Patients with PsA (classified on the basis of CASPAR criteria) and treated with blockers of the TNF-α (following EULAR recommendations) adalimumab (ADA), etanercept (ETA), golimumab (GOL), infliximab (INF) or certolizumab (CTZ), between 2011–2015, were retrospectively included. Treatment persistence was analyzed and differences between therapies were assessed by means of an analysis of variance model (ANOVA), while a chi-square test was used to evaluate relationships between therapies and causes of treatment discontinuation and the administration of concomitant disease-modifying antirheumatic drugs (DMARDs) among therapies and types of disease.

Results One hundred and ninety-six patients received: 98 ETA, 63 ADA, 22 GOL, 10 INF and 3 CTZ, with an age mean of 51.7±12.4 years old without differences among groups of therapies (p 0.993), and 111 were men (56.6%). Combination treatment with DMARDs was in 126 patients (64.3%) and no differences were observed among five anti-TNF therapies. One hundred and forty-three patients (73.0%) had persistence in first anti-TNF, mean 61.2 (12.8) months, and by groups was: 84.7% ETA, 60.3% ADA, 59.1% GOL, 70% INF and 66.7% CTZ, p 0.007, without differences among clinical subtypes (p 0.978). Combination therapy with DMARDs was associated with a higher persistence in first anti-TNF, 67.8% vs 55.9%, but without statistical significance (p 0.108). Change of first anti-TNF occurred in 53 patients: failure to treatment was in 33 patients (62.3%) and discontinuation by adverse event was in 20 patients (37.7%) without differences among groups (p 0.535). Discontinuation by failure to treatment has happen with a mean of 66.1 (43.1) months and by adverse event in 53.4 (30.2) months (p0.235). In 68 (34.7%) patients without clinical activity dosing interval of treatment was increased and we have not observed differences among the groups (p 0.265).

Conclusions In this cohort, anti-TNFα therapy was associated with high persistence at first anti-TNF therapy, mainly etanercept. Switching to another anti-TNF-α agent occurred in a quarter of patients, when treatment was suspended because of failure to response or an adverse event. Furthermore, treatment was optimized in one third of patients.

Disclosure of Interest None declared