Background Therapeutic strategy after failure of a 1st anti-TNF (TNF-IR) is not codified in rheumatoid arthritis (RA). Little is known about the criteria for choice of a second biologic by clinicians.
Objectives To determine whether patient's or disease characteristics guide clinicians for their choice of a second biologic in TNF-IR patients with RA
Methods The ROC trial randomized patients with an insufficient response to a 1st anti-TNF to receive either a second anti-TNF or a non-TNF targeted biologic. Before randomization, the clinician's preference for the second biologic (either a non-TNF targeted biologic or a second anti-TNF) was collected. After randomization, the choice of the molecule within each arm was left to the clinician.
Results The clinician's preference for the second biologic could be collected before randomization for 281 out of the 292 patients analyzed in the ROC trial. 151 clinicians would have preferred to prescribe a non-TNF targeted biologic, 72 a second anti-TNF and 58 clinicians had no preference.
Among baseline characterics of disease, only disease duration was significantly associated with the preference for non-TNF targeted biologics (median 8 [3–17] years for patients for whom clinicians would have chosen a non-TNF biologic versus 5 [1–13] years for patients for whom clinicians would have chosen a second anti-TNF, p=0.001). Other parameters (such as age, sex, body mass index, number of previous conventional DMARDs, seropositivity for RF/ACPA, extra-articular invovement, CRP, ESR, DAS28-ESR at inclusion) has no significant impact on the clinician's choice.Cotreatment either with glucocorticoids, methotrexate or anopther conventional DMARD had no significant impact on the choice of clinicians either.
With regards to the choice of the mechanism of action in the non-TNF targeted group, clinicians significantly more frequently prescribed abatacept in patients with cardiovascular risk factors (70% of patients treated with abatacept had cardiovascular risk factors compared to 44% of patients treated with rituximab or tocilizumab) p=0.03). Clinicians prescribed significantly more frequently tocilizumab in patients with a high C-reactive protein value (median baseline value of 10.5 mg/l versus 5 mg/l in patients who received abatacept or rituximab) (p=0.01).
Conclusions There is no characteristic patient's profile explaining the preference of clincians for a non-TNF biologic or a second anti-TNF in TNF-IR patients. Among non-TNF biologics, abatacept was more frequently prescribed in patients with cardiovascular risk factors and tocilizumab was more frequently prescribed in patients with a high C-reactive protein value.
Disclosure of Interest None declared