Tapering and discontinuation of glucocorticoids in patients with rheumatoid arthritis treated with tofacitinib

This study aimed to evaluate the steroid-sparing effect of tofacitinib in patients with RA. To the best of our knowledge, this is the first prospective study that evaluated the feasibility of a predetermined schedule for GCs discontinuation in patients initiating tofacitinib treatment. This showed that already 12 weeks after starting treatment, 30% of patients were able to stop their daily dose of GCs. The most recent recommendations for the treatment of RA suggest that the dosage of GCs should be reduced to discontinuation as soon as possible¹. In fact, the dose and duration of steroid treatment influence the safety profile and there is no agreement on the definition of the ‘safe’ dose.¹⁷. The optimal dose and duration of GC therapy and the best strategy to taper off and withdraw GCs are still scarce. A recent systematic literature review reported 14 different regimens proposed to reduce and/or stop oral GCs¹⁸. Randomized clinical trials with RA drugs do not provide sufficient information on GC therapy during treatment with biologics and do not allow to displace definitive conclusions on GC discontinuation, also considering that the steroid-sparing effect never appears as a secondary endpoint¹⁹.

A subanalysis of the ACTION (AbataCepT In rOutiNe Clinical Practice) study showed that over a 24-month period from the start of abatacept, approximately 40% of the 734 patients taking GCs at the start of the study were able to control their prednisone levels to lower. dose, usually within the first 3 months of treatment, and that the median GC dose decreased from 7.5 mg/day to 5 mg/day after 24 months of treatment²⁰. The need to evaluate the steroid-sparing effect of RA medications is real and relevant, as evidenced by real-world studies. Most real-world evidence comes from retrospective studies evaluating the effect of bDMARDs on GC therapy in patients with long-standing RA, such as those enrolled in our study. Many years ago, Naumann et al. described dose reduction of GCs that was associated with reduced disease activity as early as 3 months after starting a TNF inhibitor in 87 patients with RA; during 5 years of follow-up, 81% of patients reduced the dose of GCs and 32% of patients discontinued treatment²¹.

In a French retrospective study published in 2009, the authors noted an overall 30% reduction in the dose of oral GCs in the first year of treatment with TNF inhibitors, starting as early as 3 months; on the other hand, 61% of patients were still using low-dose prednisone after 1 year⁷. More recently, analysis of the TReasure database found that after a median of 59 months, 28.4% of 1936 patients receiving GCs at registry entry discontinued concomitant steroid treatment.²². The retrospective nature of the study explains the lack of prespecified criteria for GC tapering. The SPARE-1 ​​study included RA patients treated with tocilizumab (TCZ) and an oral prednisone dose > 5 mg/day and aimed to evaluate the percentage of patients who reduced their GC dose to less than 5 mg/day after 12 months of TCZ²². At the end of follow-up, 40% of patients reached the target dose; RA duration not more than 5 years, daily prednisone dose <7.5 mg, and low baseline ESR were predictive of PDN reduction to less than 5 mg/day²³.

Fernandez-Nebro et al. prospectively followed for 24 months 161 patients starting infliximab, etanercept, or adalimumab, showing that almost 60% of patients taking GCs at baseline were able to discontinue and the remaining patients significantly reduced the daily dose²⁴. In a previous paper published by our group, we demonstrated a 56% reduction in the percentage of RA patients treated with baricitinib who were taking glucocorticoids after 24 weeks; In addition, we registered a significant reduction in the daily dose of prednisone from an average of 5 mg/day to 0 mg/day, simultaneously with a significant reduction in pain after just 4 weeks.²⁵. Similarly, in the current study, the rapid effect of tofacitinib on pain and disease activity certainly contributed to the rapid reduction of the dose of GCs, to complete discontinuation; Indeed, approximately one-third of included patients discontinued daily GCs within 12 weeks of starting tofacitinib.

The initial adjunctive GCs therapy on a DMARD background would allow rapid control of disease activity, also guaranteeing long-term structural benefits, which persist even after GCs withdrawal; doses ranging from 5 to 10 prednisone per day appear to be associated with significantly slower radiographic progression². A post hoc analysis of 6 randomized, controlled phase III trials of tofacitinib suggested that concomitant use of GCs did not affect either the clinical or radiographic efficacy of the drug.²⁶. In a prospective evaluation of the efficacy of tofacitinib according to discontinuation of GCs, we observed no difference in the reduction of disease activity as assessed by DAS28_CRP, CDAI and SDAI; furthermore, low disease activity or remission was achieved regardless of the concomitant use of GCs. In contrast, data from a Turkish registry showed a significantly higher DAS28_CRP score and a significantly lower proportion of patients achieving low disease activity or remission in patients who continued GCs concurrently with bDMARDs or tsDMARDs.²⁰. Data on the effect of GCs on clinical response to TNF inhibitors are also conflicting and have shown that a higher or lower percentage of patients achieve the treatment goal.^27,28. In the SEMIRA (The Steroid EliMination In Rheumatoid Arthritis) trial, patients were randomly assigned to the regimen of continued prednisone or tapered prednisone; two-thirds of patients who achieved low disease activity with tocilizumab tapered the steroid dose in the 24-week study, but contrary to what we observed in our patients, patients who continued with prednisone had better control of disease activity²⁹.

In our study, age at enrollment and disease duration were negatively associated with GC discontinuation. This is not surprising, as results from a recent double-blind, placebo-controlled study showed that elderly patients (over 70 years at baseline) with established and severe RA had long-term beneficial effects in both disease and disease activity. and joint damage resulting from taking a low dose of prednisolone as an addition to basic treatment³⁰. However, this study did not include patients taking tsDMARDs. Recently, data from the Veterans Affairs Rheumatoid Arthritis Registry showed that 54% of patients tapered and 33% discontinued oral GCs; younger age, positive rheumatoid factor, higher ESR at enrollment, a greater number of prior csDMARDs, and a higher mean glucocorticoid dose during the 30 days before the index date were all significantly associated with GC tapering and discontinuation³¹.

The 2013 EULAR recommendations on the treatment of moderate to high doses of glucocorticoids in rheumatic diseases state that “When deciding to initiate glucocorticoid treatment, comorbidities and risk factors for adverse effects should be evaluated and treated where indicated […]”³². Particularly at medium to high doses, GCs are often associated with the occurrence of adverse events, including osteoporosis, diabetes, hypertension, cardiovascular events, and recurrent infections. Sparing GCs – as suggested by our results – may be beneficial in terms of short-term safety (risk of infection, especially reactivation of herpes Zoster) and long-term safety (metabolic and cardiovascular events, as well as osteoporotic fractures ). A large study of RA patients from the CorEvitas registry showed that GCs increased the risk of cardiovascular events after just six months of treatment, at a daily dose of more than 5 mg³³.

Some of these side effects – mainly infections and cardio-metabolic effects – raise concerns in patients treated with both bDMARDs and tsDMARDs; in particular, the results of the ORAL Surveillance study raised a warning about cardiovascular safety in patients treated with tofacitinib³⁴.

The current study has some limitations. The small sample size did not allow stratification of patients based on concurrently administered MTX. However, the MTX dose remained stable and the logistic regression did not identify the co-medication as a factor influencing the ability to discontinue GCs. The long disease duration and failure of previous treatments probably caused the steroid-sparing effect of tofacitinib to be underestimated. Still, 30% of patients stopping GCs after just three months is an impressive result.

The strength of the study lies in its prospective nature and interventional design, which allowed the GC dose reduction schedule to be predefined to homogenize the outcome.

In conclusion, in our cohort of patients with long-standing RA treated with tofacitinib, discontinuation of glucocorticoids was feasible in up to 30% of patients. The results of the study should encourage rheumatologists to consider GC tapering and withdrawal as a possible goal in the daily management of RA patients on long-term treatment with oral GC.