Comparative effectiveness of treatment with the first TNF antagonist in monotherapy, the first TNF antagonist plus one conventional synthetic disease-modifying antirheumatic drug, and the first TNF antagonist plus two or more conventional synthetic disease-modifying antirheumatic drugs in patients with rheumatoid arthritis

Abstract

Background

Rheumatoid arthritis (RA) patients are treated with a mean of 3-4 conventional synthetic disease-modifying antirheumatic drugs (csDMARDs) with or without glucocorticoids (GCs), before the first biologic prescription. The main reasons for change are inefficacy in 30-40 % of patients, and toxicity [asymptotically =] 10 %. Thus, they are treated with the first TNF antagonists in monotherapy. The aim of this study was to analyse the csDMARD and GC prescription patterns before and during treatment with the first TNF antagonist, and compare their effectiveness in three groups of patients.

Methods

An observational, prospective, multicentre study in common clinical practice was designed. Treating rheumatologists recorded patient variables, including previous and concomitant csDMARDs and GCs in a database. The data were analysed using descriptive, inferential and multivariate statistics.

Results

There were 1136 patients included; 21 % received the first TNF antagonist in monotherapy, 67 % received the first TNF antagonist plus one csDMARD, and 12 % the first TNF antagonist plus two or more csDMARDs. Most patients were female (73 %), RF+, and ACPA+, and had erosions; mean age was 53.2 (±13.0) years, and duration of disease was 9.1 (±7.6) years. They had high activity with DAS28 of 5.8 ± 1.1, and poor physical function with HAQ of 1.43 ± 0.63, and significant differences between groups in clinical variables and comorbidities; 94 % had received treatment with GCs, MTX, LFN, or SSZ at any time before the first TNF antagonist, 5 % (n = 52) had been treated with CLQ or HCLQ, and 1 % (n = 13) had received neither GCs nor csDMARDs. Before the first TNF antagonist, the drugs most commonly used were GCs (78 %), MTX (50 %), LFN (44 %), and SSZ (21 %). Concomitantly with the first TNF antagonist, 977 patients (85 %) were receiving GCs, MTX, LFN, or SSZ; 15 % (n = 173) received their first TNF antagonist without any concomitant GCs or csDMARDs, true monotherapy, and 6 % received their first TNF antagonist with GCs. The drug most commonly used at the time of first TNF antagonist initiation was MTX (58 %). All treatment groups had clinically and statistically significant improvements in DAS and HAQ scores. Effectiveness analysis (controlling for confounders) showed mean drug survival of 16.7, 20.1 and 11.7 months in each group, respectively (p < 0.001). The model that best explained a good EULAR response included the baseline and 6-month DAS28.

Conclusions

The three groups of patiernts, have different comorbidities and disease characteristics. Treatment with low or very low doses of GCs is common. True monotherapy with the first TNF antagonist without prednisone or csDMARDs is infrequent. After controlling for potential confounders, effectiveness was a little different.