Background/Purpose:

Previous studies have shown differences in treatment retention between anti-TNF agents. Furthermore, although inconsistent, data from the literature suggest that some factors (e.g. concomitant methotrexate use) may be associated with improved retention of certain anti-TNF agents. The aim of this analysis was to compare the long-term retention of infliximab (IFX) and golimumab (GLM) and to identify independent predictors of retention in patients with rheumatoid arthritis (RA) followed in Canadian routine practice 

Methods:

BioTRAC is an ongoing, prospective registry of patients initiating treatment for RA, ankylosing spondylitis, or psoriatic arthritis with IFX or GLM. Eligible people for this analysis included RA patients treated with IFX who were enrolled since 2002 or with GLM enrolled since 2010; patients not discontinued who had a follow-up of <18 months were excluded. Independent predictors of long-term retention, defined as ≥18 months, were assessed with multivariate logistic regression using backwards variable selection. Receiver operator curve (ROC) analysis was used to determine the optimal cut-off points of CDAI for long-term retention

Results:

A total of 972 patients (76.3% female) were included with mean (SD) age of 55.9 (13.8) years and disease duration of 8.6 (9.1) years at baseline. The majority were biologic naïve (94.5%), treated with IFX (84.8%), and received a concomitant DMARD (88.5%) at baseline; 35.9% were treated with a corticosteroid (CS). Mean (SD) disease parameters at baseline were: CDAI: 35.3 (17.5); HAQ: 1.6 (0.7); swollen joint count (0-28): 10.3 (6.8); tender joint count (0-28): 12.0 (7.9); patient global (VAS cm): 6.0 (2.5); physician global (0-10 NRS): 6.4 (2.1).

Six hundred (61.7%) patients received treatment for ≥18 months. In univariate analysis, age [OR (95% CI): 0.99 (0.98-1.00)], disease duration [0.97 (0.95-0.99)], enrolment year [1.10 (1.06-1.15)], male gender [0.68 (0.50-0.94)], GLM vs. IFX use [2.24 (1.58-3.19)], biologic naiveté [0.58 (0.33-1.01)], and baseline DMARD use [0.54 (0.36-0.80)] were identified as potential predictors of retention (P<0.100). Baseline CS use [0.86 (0.66-1.13)], CDAI [1.00 (0.99-1.00)], and HAQ [0.98 (0.81-1.17)] did not have an impact. Upon multivariate adjustment, older age [0.98 (0.97-1.00)] and male gender [0.50 (0.29-0.85)], were associated with significantly lower odds of discontinuation, while more recent enrolment year was associated with higher odds [1.36 (1.20-1.54)].

Lower disease activity (CDAI remission vs. low vs. moderate vs. high) at 6 or 12 months was associated with significantly (P<0.001) higher probability of long-term retention. ROC curve analysis identified a CDAI score at 6 months ≤15.7 (AUC=0.652; P<0.001) and a score at 12 months ≤14.6 (AUC=0.679; P<0.001) as most accurately predicting long-term retention  

Conclusion:

The results of this analysis have shown that treatment retention of IFX and GLM in real-world are comparable. Age, gender, and enrolment year, possibly signifying differences in patient preferences or changes over time in clinical practice, as well as CDAI score of 15 or less at 6 or 12 months were identified as significant independent predictors of long-term retention

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/predictors-of-long-term-retention-of-infliximab-and-golimumab-in-rheumatoid-arthritis-an-analysis-from-a-prospective-observational-registry/