Background Treat to target strategy trials have been shown to improve patient outcomes in the research setting. However, a general goal in exploring targeted therapy in rheumatoid arthritis (RA) will continue to be maximizing flexibility of the rheumatologist in treatment decisions. The Objectives Study in Rheumatoid Arthritis (OSRA) was a targeted biomarker randomized controlled trial (RCT) that compared two different biomarkers head to head against a usual care control arm.
Objectives To evaluate which clinical, laboratory and patient self-report measures drive rheumatologists’ treatment decisions in the “usual-care” non-target control arm.
Methods Active RA patients were randomized to 1 of 2 treat-to-target arms: swollen joint count (SJC) <3 or normal CRP; or to a usual care arm. Patients were seen monthly for 24 months by a study nurse. The study nurse collected data on drug treatment, patient self report clinical status, joint count, laboratory tests and patient safety. The treatment and the interval between visits of patients randomized to the usual care were determined by the treating rheumatologist.
Results Of the 249 patients recruited by 31 rheumatologists, 82 were randomized to usual care, given a total of 1817 research patient-months. Mean age and disease duration were 56 and 7 years. Baseline DAS28 (ESR) was 5.08 (1.6-7.95). Patients were seen by their rheumatologist usually every 2 to 4 months. Treatment as a percentage of study nurse patient-months was: methotrexate 79%, prednisone 61%, leflunomide 31%, hydroxychloroquine, 30%, sulphasalazine 29%. Only 10% were treated with a TNFalpha blocker. Use a 5-point treatment change scale, on average, DMARD treatment was reduced in 222 visits, unchanged in 1314 visits and increased in 255 visits. The average DAS28 was 4.11 for treatment reduction, 3.95 when treatment was unchanged, 5.16 when treatment was increased and 6.03 with introduction of a TNF blocker. These differences were significant (p<0.02). Other measures of RA activity, no change versus increase in DMARDs, were as follows: 7 vs.11; 6 vs.10; 20 vs.32; 10 vs.23; 23 vs.38; 24 vs.39; 0.82 vs.1.2 for SJC, TJC, ESR, CRP, patient global, patient pain, and HAQ, respectively. In a fixed-effects regression the only components of the DAS (i.e. SJC, SJC, ESR, patient global) that significantly predicted treatment decisions were SJC and ESR; with a explanatory power of only 11%.
Conclusions In an analysis of the usual care arm from a biomarker-target randomized controlled trial of patients with active RA, implicit rheumatologists’ treatment decisions are not clearly explained by standard measures of RA disease activity. Furthermore, DAS scores of moderate disease activity, on average, did not lead to increases in DMARD treatment. Patient preferences, including patient adaptation, or toxicity/adverse effects may explain some of these findings, which in turn, may impact treat to target strategies. However, with increased use, experience and breadth of effective biologic agents, rheumatologists may be more likely to engage in treat to target strategies.
Disclosure of Interest M. Lassere Grant/Research support from: Unrestricted Educational Grant Sanofi Aventis, J. Edmonds Grant/Research support from: Unrestricted Educational Grant Sanofi Aventis, K. Johnson: None Declared