Will we ever agree on using low-dose glucocorticoids in treating rheumatoid arthritis?

This editorial refers to ‘Safety and efficacy associated with long-term low-dose glucocorticoids in rheumatoid arthritis: a systematic review and meta-analysis’, by Palmowski et al. 2023;62:2652–60.

Discussion on an acceptable balance between efficacy and safety of low-dose glucocorticoids (GCs) in the treatment of RA is as old as the treatment itself. Most rheumatologists have a clear opinion on the use of GCs in RA; either they use them with caution (‘conscious users’) or they are trying to avoid their use (‘try to avoiders’). The same holds for patients, although in general they are more positive about the use of GCs since they experience clear beneficial effects [1].

The systematic literature review (SLR) published in this issue adds new information to this discussion [2]. Surprisingly for some, the conclusion of their analysis is that long-term low-dose GCs have a clear beneficial effect on disease activity and disease progression without an increase in (severe) adverse events ((S)AEs), except for an increase in infections in 40% of patients.

The efficacy of GCs has been clear from the first patient treated, a young lady who rose from her bed, like Lazarus in the bible. Dosages used at that time were high, and soon AEs became apparent and led to more cautious use of GCs. Since then, hundreds of articles on the use of GCs in rheumatic and other inflammatory diseases have been published, without reaching a consensus for optimal clinical use. Why is it so difficult to gather scientific evidence to reach a consensus between ‘conscious users’ and ‘try to avoiders’? The most important factor is bias by indication: those with the most active, worst RA are most often treated with GCs. These are exactly the patients with the most AEs and most comorbidities, irrespective of their treatment with GCs. Also, personal experience has a major influence: nearly every rheumatologist has a memory of an (S)AE in a patient that needed to be treated with GCs, influencing her/his attitude for years to come.

To bring ‘conscious users’ and ‘try to avoiders’ more together the EULAR Task Force on GCs organized a consensus meeting that came to an interesting conclusion: long-term use of GCs in daily dosages of ≤5 mg leads in general to more benefits than harm, while long-term use of GCs in dosages of ≥10 mg leads in general to more harm than benefits. The balance for daily dosages between 5 and 10 mg depends on specific patient characteristics and risk factors. As an example, in case of diabetes mellitus, risk increases with genetic disposition, age, obesity and chronic inflammation, while risk decreases with weight loss when obese, a healthy diet and appropriate exercise [3]. It should be noted that in the field of osteoporosis, 5 mg of prednisone daily is slowly gaining acceptance as well [4].

To underpin the update of the EULAR recommendations on the treatment of RA an SLR on the use of GCs was published recently [5]. The authors rightfully considered that randomized controlled trials (RCTs) are generally not powered to evaluate safety outcomes, that they often exclude patients at high risk of AEs and that most have a relatively short follow-up. Therefore, they considered observational studies most relevant to evaluate safety. Thus they accepted confounding by indication, and this bias influenced (again) interpretation of their results on safety: ‘This SLR confirms well-known safety risks of GC use such as an increased risk of osteoporotic fractures, serious infections, diabetes, cardiovascular events and mortality, but with large heterogeneity between studies in terms of definitions of outcome measures and exposure categories, and particularly methods to cope with bias’ [5].

To avoid bias by indication, the Glucocorticoid LOw-dose in RheumatoId Arthritis (GLORIA; NCT02585258) study was initiated, a long-term (2 years) placebo-controlled RCT in a large number of patients, thus allowing a better evaluation of safety [6]. Inclusion and exclusion criteria for this RCT are unusual: the researchers wanted as few limitations as possible, so only unstable cardiovascular diseases and unstable diabetes were excluded, and only elderly people (>65 years of age) were included. Also, all other medications (except for actual or recent use of GCs) were allowed. These criteria led to participants with a more prolonged disease, and quite often comorbidities. In fact, this population was enriched for AEs. More than 450 patients from seven different European countries used 5 mg of prednisone or placebo daily for 2 years, in addition to their DMARD. The well-known beneficial effects of GCs on disease symptoms and disease progression were confirmed and, remarkably for many, no increase in (S)AEs was found.

To put these findings on a comparable number of AEs in broader perspective, authors pooled their data with existing literature, the SLR published in this issue [2], to estimate the ‘unbiased’ incidence of AEs due to GCs. Six RCTs were identified that lasted for at least 2 years and compared low-dose GCs (≤7.5 mg daily) with placebo. Use of low-dose GCs for 2 years did not lead to an increase in AEs, SAEs, withdrawal due to AEs or to an increase in ‘well-known’ AEs of GCs such as osteoporosis, fractures, cardiovascular diseases, diabetes and hypertension. Only an increase in infections was noted [relative risk 1.40 (95% CI 1.19, 1.65)], which was of the same order of magnitude as biologic DMARDs (bDMARDs). The authors concluded that in most patients with RA, long-term low-dose GC use has an acceptable safety profile. The SLR and meta-analysis are well performed, confirming most recent insights in this field. Of course, there are some caveats: two of the older RCTs had some risk of bias, as they might be less observant for AEs, although serious AEs were well noted. Therefore, this study can be considered an important argument for an acceptable benefit:risk ratio for the use of long-term low-dose GC therapy in RA.

What about slightly higher dosages? A study that was not included in this SLR, because of a higher dosage of prednisone, also meticulously reported AEs [7]. That study looked at the effect of 10 mg prednisone daily for 2 years vs placebo in a methotrexate-based tight control study in >230 patients and reported a marginally decreased number of AEs: 86 in the prednisone group and 94 in the placebo group. This decrease in AEs was especially noted in liver toxicity and nausea, and the authors speculated that significantly lower use of NSAIDs in the prednisone group (40% of the placebo) could explain part of this difference [6]. Sparing of NSAIDs might indeed be an interesting by-effect of using low-dose GCs. A 10-year follow-up of this study showed a lower initiation rate of the first bDMARD and significantly better radiographic outcomes, without resulting in more GC-related comorbidities [8].

The present SLR on AEs due to long-term low-dose GCs may give us more trust in the conscious use of GCs; the evidence is clearly accumulating, but it is doubtful whether everybody will be convinced [9]. We will continue to have ‘cautious users’ vs ‘trying to avoiders’, but the view of some might change a bit. Whatever changes are made, GCs will continue to be an important part of our toolbox for the treatment of patients with RA for many years to come.

Data availability

No new data were generated or analyzed in support of this article.

Funding

No specific funding was received from any bodies in the public, commercial or not-for-profit sectors to carry out the work described in this article.

Disclosure statement: In the past, the author was the coordinator of the EULAR Glucocorticoid Task Force.

References