Background For rheumatoid arthritis (RA) patients who discontinue the first biologic agent (bDMARD), most commonly being a TNF inhibitor (TNFi), there is little evidence supporting the next best choice between a second TNFi course or a non-TNFi bDMARD in clinical practice.
Objectives To compare the effectiveness and the adherence to therapy with non-TNFi versus TNFi administered as the second-line bDMARD in RA patients with one prior TNFi use.
Methods All patients starting a bDMARD in the Rheumatology Department of the University Hospital of Heraklion, Crete, are included in a prospective observational study after their written informed consent. Data concerning disease activity at pre-specified time-points, drugs, comorbidities and any adverse events are recorded. For the present study we analyzed patients with RA starting their second course of a bDMARD after discontinuation of a TNFi. We compared DAS28 difference at 6 and 12 months using linear regression analysis and treatment retention using Kaplan-Meier survival curves with log-rank test.
Results A total of 384 patients started a 2nd (different) TNFi [N=213 (Infliximab: 26, Adalimumab: 77, Etanercept: 89, Golimumab: 13 and Certolizumab: 8)] or a non-TNFi [N=171 (Rituximab:71, Abatacept:66, Tocilizumab:34)]. Patients' characteristics at baseline are described in the Table.
Two-year drug survival was higher for non-TNFi (64% vs. 39%, log rank p<0.001) due to lower frequency of discontinuations for primary failure (p<0.001) and adverse events (p=0.019).
δDAS28 was comparable between non-TNFi and TNFi patient groups both at 6 [mean (SD): -1.16 (1.29) and -1.07 (1.55) respectively, p=0.296] and at 12 months [-1.41 (1.29) and -1.39 (1.26) respectively, p=0.670]. In patients who did not receive co-therapy with methotrexate, significantly greater δDAS28 was observed with a non-TNFi (-1.25 (1.29) vs.-0.68 (1.61), p=0.006). When the first TNFi was discontinued due to primary failure, we observed a trend for greater δDAS28 in the non-TNFi patient group compared to the 2nd TNFi group (-1.4 vs. -1.0, p=0.12) while the opposite was observed in patients who have experienced secondary failure to the 1st TNFi (-0.81 vs -1.48, p=0.18), but this did not reach statistical significance, probably due to the low number of available patients.
Conclusions In RA patients who need a 2nd bDMARD after discontinuation of a TNFi, administration of a non-TNFi results in similar clinical responses but higher treatment adherence compared to a second TNFi agent. In patients who do not receive methotrexate, responses are better with a non-TNFi bDMARD.
Disclosure of Interest None declared