With several new biologic (b) disease-modifying antirheumatic drugs (DMARD) and targeted synthetic (ts) DMARD now available for the treatment of rheumatoid arthritis (RA), information regarding their comparative effectiveness would be of interest. Relatively few head-to-head trials have been published, however. Though "real-world" studies have been published to provide some information about comparative effectiveness, Deakin and colleagues used a target trial emulation framework to apply clinical trial methods to real-world data. Using the Australian OPAL registry of bDMARD/tsDMARD-naive patients, they developed a randomized controlled trial protocol of tofacitinib vs adalimumab using an intention-to-treat analysis. Under this framework, there was small reduction of disease activity with tofacitinib vs adalimumab at 3 months and no difference at 9 months. While this framework is conceptually interesting, it may be more meaningful used in side-by-side comparison to a real-world analysis of the same data to evaluate pitfalls and biases in both; otherwise, its utility as a stand-alone analysis of observational data is not fully clear.
Østergaard and colleagues also performed a head-to-head study of several different therapies to address the question of optimal treatment strategies for patients with early RA. Patients with moderate to severe disease activity were randomly assigned to treatment with methotrexate combined with (1) oral glucocorticoid or sulfasalazine, hydroxychloroquine, and intra-articular steroid injections, (2) certolizumab, (3) abatacept, or (4) tocilizumab. Disease activity and radiographic changes were evaluated at 48 weeks. In this study of over 800 patients, treatment with abatacept or certolizumab was associated with improved Clinical Disease Activity Index (CDAI) remission rates compared with the active conventional therapy (group 1), but tocilizumab was not. The overall differences between bDMARD treatment groups were small and thus may not reflect significant differences in effectiveness. Instead, this study challenges the notion of initiating conventional synthetic DMARD (csDMARD) therapy in patients with early RA and stepping up to bDMARD, as initial bDMARD therapy may be of benefit in patients with more active early RA.
Alongside the question of the effectiveness of bMARD and tsDMARD in real-world settings, the appropriate role for long-term low-dose prednisone in the treatment of RA remains unknown. A recent study by Güler-Yüksel and colleagues examined the effects of 5 mg prednisolone daily in addition to standard therapy in patients over 65 years of age with active RA. Due to the potential complications of weight gain and glucose intolerance with long-term glucocorticoids, in addition to low-bone-density issues, their use has generally not been viewed favorably. In this multicenter trial, 449 patients were randomly assigned to receive prednisolone vs placebo in addition to their usual medications over 2 years. Notably, patients in the prednisolone group had an average of 0.9 kg weight gain compared with placebo with 0.4 kg weight loss over 2 years. By the end of 2 years, 29% of patients in the prednisolone group had a weight gain of > 2 kg compared with 18% of patients in the placebo group. Only 57 patients in all underwent body composition analysis, and, interestingly, those in the prednisolone group had small increases in lean body mass compared with fat mass, though these patients were not necessarily representative. The authors suggest, though the study does not prove, that low-dose prednisolone can be protective against sarcopenia, which is associated with older age and "rheumatoid cachexia." The study also did not examine the interaction of glucocorticoid use with diet and exercise. While it is reassuring that patients in this study did not experience major weight gain, it does not appear to be a generalizable finding at this point.