Early response as a predictor of long-term clinical response in DMARD-naïve patients with severe, active and progressive RA treated with certolizumab pegol plus optimized MTX versus optimized MTX alone

Background: C-EARLY (NCT01519791) assessed the efficacy and safety of certolizumab pegol (CZP)+optimized MTX vs placebo (PBO)+optimized MTX in inducing and maintaining a sustained clinical response, and inhibiting radiographic damage, in DMARD-naïve patients (pts) with active, severe, progressive RA.1 Here, we examine the association between an inadequate response to PBO+MTX or CZP+MTX treatments at Week (Wk) 12 (lack of improvement from baseline [BL] in DAS28[ESR], DAS28[CRP] or CDAI) and corresponding response at Wk 52. Objectives: To examine the association between Wk12 response and Wk52 clinical response (remission [REM]/low disease activity [LDA]) in recently diagnosed DMARD-naïve pts with active, severe, progressive RA, treated with CZP+optimized MTX vs PBO+optimized MTX. Methods: This multicenter, double-blind, randomized controlled trial enrolled pts who were DMARD-naïve with active, severe, progressive RA (<1yr since diagnosis, fulfilling 2010 ACR/EULAR criteria). 879 pts were randomized 3:1 to CZP (200mg Q2W+MTX; n=660) or PBO Q2W+MTX (n=219). MTX was initiated at 10mg/wk and increased to 25mg/wk by Wk8; maximum tolerated dose per pt (mean optimized dose

, CZP:

21mg/wk

, PBO:

22mg/wk) was maintained to Wk52. Negative predictive values (NPV) were defined as probability of failing to achieve Wk52 REM (DAS28[ESR]/[CRP]<2.6 or CDAI≤2.8) or LDA (DAS28[ESR]/[CRP]≤3.2 or CDAI≤10) after failing to achieve a corresponding defined response at Wk12; change from BL in DAS28(ESR)/(CRP)≥0.6 or ≥1.2/CDAI≥6 and ≥12. DAS28(CRP) LDA/REM definitions are not validated. Observed data were utilized for Wk12 responses; missing Wk52 values were imputed using non-responder imputation. Results: The NPV of Wk12 responses were high (Table): PBO+MTX pts who did not achieve DAS28(ESR) improvements from BL ≥0.6 had 94% probability of not being in REM or LDA at Wk52. CZP+MTX pts who did not achieve DAS28(ESR) improvements from BL ≥1.2 had 93% probability of not being in REM at Wk52. NPV of changes in DAS28(CRP) and CDAI at Wk12 were less informative for predicting remission at 1yr (Table). Conclusions: In early RA pts treated with optimized-dose MTX, failure to achieve a ≥0.6 DAS28(ESR) reduction from BL at Wk12 was associated with a low probability (6%) of DAS28(ESR) REM/LDA at Wk52, supporting treatment step-up at 3 months. For pts with active, severe RA with poor prognostic factors where CZP has been initiated as first-line treatment, failure to achieve a rapid decrease of 1.2 points was associated with a reduced likelihood of achieving REM (7%). These findings in DMARD-naïve pts are consistent with the negative predictability outcomes for CZP in established RA.2. (Table Presented).