Background/Purpose:Anti-TNF has drastically changed the prognostic of the RA,carrying an important healthcare expense.This is why,optimisation seems a successful strategy that should not be linked to a worse become of our patientsÕ clinical evolution.

Methods: Observational study of the prospective cohort RA-Paz. Aim of the study:describe a population of patients with RA under optimised treatment with Adalimumab(Ada) or Infliximab(Ifx).Study the incidence of flares and establish predictive factors of flares at baseline and pre-optimization.All the patients diagnosed of RA under treatment with Ifx or Ada between Jan.2000 and Dec.2016 of the day-care unit,were included.Demographic data,clinical activity and blood sample results were collected at baseline,pre-optimization(pre-op)and at 3,6,9,12,18 and 24 months.Drug serum trough levels were measured under ELISA in each visit.Optimal range for Ifx was described as drug concentration between 1000- 4000 ng/ml and 1500- 5000 ng/ml for Ada.Optimisation was defined as drug use below standard dose.Flares were collected from the pre-op visit.Flare was described as clinical worsening which led to a therapeutic change or a DAS28>3.2 and DeltaDAS28>0,6.Predictive factors of flare at baseline and pre-op were evaluated with a uni and multivariate analysis.

Results: Of the 271 patients diagnosed of RA, 74 patients were optimised (44 under Ada and 30 under Ifx). During the 24 months after the pre-op visit,55,4%(41)of the patients presented at least one flare,with an average of 1,38 flares[1-5]. 53.7% of the patients,(22/41),were controlled with the adjustment of non biological treatment.Only 39,0%(16) of the patients,had to go back to the previous optimised dose and 7,3%(3)to the standard dose.88%(39/41)were controlled after the dose modification.104 flares were collected,33%(34) happened at the 3rd month and 20%(21) at 24th. In the population who presented flares, we observed a persistent higher DAS Vs the patients who never presented flares(DAS pre-op 3,20±1,16 Vs 2,26±0,59;DAS 24month 3,61±1,13Vs2,10±0,65;p=0,007). A least proportion of patients with flares were in supra-optimal range(13,3% with flares vs 26% without, p=0,007). At baseline, no clinical factors were predictive of flare. Nor were blood sample results. In contrast, a higher disease activity, measured by DAS pre-op(p=0,004),a worst EULAR answer(p=0,027) and not being in supra-optimal range(p=0,032),were statistically correlated with flares development at the univariate analysis. Time to the optimisation tended to the significance(OR=1,152; p=0.08). In the multivariate analysis, only a higher DAS pre-op(OR:2,00,[1,08-3,73])and being in optimal(OR:5,90, [1,38-25,2])and sub-optimal range(OR=6,05[1,28-28,7]), were independently correlated.

Conclusion: We noted a high proportion of flares in our cohort of optimised patietns. However, flares were controlled with dosage readjustment without needing a chanfe of the treatment. Independently correlated predictive factors for flares were a higher disease activity measured by DAS and not being in therapeutic range in the pre-optimisation visit.