Background Since approval of the first biologic (b) DMARDs for the treatment of rheumatoid arthritis (RA) in 2001, bDMARDs have been utilized to treat patients with severe disease. However, characteristics of patients treated with bDMARDs in the first years after approval differ substantially from those receiving bDMARDs nowadays.
Objectives To investigate which clinical parameters influence the decision of rheumatologists to start the 1st bDMARD treatment in patients with RA over the past 15 years.
Methods We used data of a prospectively observed cohort of RA patients from the German biologics register RABBIT. Since 2001, patients have been enrolled at start of a conventional synthetic (cs)DMARD or bDMARD after ≥1 csDMARD failure. 13,568 RA patients were included until April 2015. After exclusion of patients with prior biologic treatment (n=2558), insufficient information on prior treatment (n=579) or inclusion in 2007/2008 (n=918), where no patients with csDMARD start were recruited, 9.513 patients remained for the analyses. They were stratified according to their year of inclusion in RABBIT (2001–2003, 2004–2006, and 2009–2015). We applied a machine learning method of model-based boosting to select clinical parameters which have a relevant impact on treatment decisions in each of the three episodes and compared the results with multiple logistic regression models.
Results N=1781 biologic-naive patients were included during 2001–3, n=2781 in 2004–6 and n=4951 after 2008. The proportion of patients included with the start of a bDMARD was consistently between 60–62%. Patient characteristics at start of the first bDMARD changed over time. Mean disease duration decreased from 11.9 to 9.1 years, mean number of DMARD failures declined from 3.6 to 2.2 and mean disease activity (DAS28) fell from 6.0 to 5.0. The proportion of patients with osteoporosis declined from 24.5% to 11.9%.
High disease activity (DAS28), prior DMARD failures and high doses of glucocorticoids (GC) were associated with starting a biologic treatment independent from the enrollment year. Over time, more patients with only moderate disease activity, lower doses of GC and positive rheumatoid factor as well as those with more severe comorbidities like heart failure and chronic renal disease were included in the bDMARD cohort. Male sex was associated with the start of bDMARD in 2001–2003 but the effect receded in recent years of enrollment.
Conclusions The increasing use of biologics in patients with moderate disease activity and GC treatment shows that new treatment guidelines recommending earlier use during disease course to prevent further damage are realized in daily care practice. Growing knowledge and experience with biologics increasingly allows rheumatologists to treat also patients with serious comorbidities like heart failure.
Disclosure RABBIT is supported by a joint, unconditional grant from AbbVie, Bristol-Myers Squibb, Celltrion, Hospira, MSD Sharp & Dohme, Pfizer, Roche, and UCB.
Disclosure of Interest None declared