Taper, Dose Down, Discontinue: Striving for Less Steroid Use

Now, 75 years after the first presentations were made on the “sensational” effects of cortisone in the treatment of rheumatoid arthritis (RA), glucocorticoids (GCs) are still highly relevant and widely used in the management of RA and other immune-mediated inflammatory diseases.
“It makes me smile because this is such an old drug, and we need it still so much. It still hasn’t been replaced,” Josef Smolen, MD, observed during the European Alliance of Associations for Rheumatology (EULAR) 2024 Annual Meeting.
At low doses, GCs are highly effective as anti-inflammatory and anti-destructive agents in RA and many other diseases, said Smolen, a rheumatologist and immunologist and professor emeritus at the Medical University of Vienna, Vienna, Austria.
But even after all this time, the mechanisms that lead to efficacy vs toxicity have yet to be clarified. “Such separation may provide further insights into future treatment options,” said Smolen.
His comments, made during a special session on the 75th anniversary of GCs at EULAR 2024, underscore the endless saga to manage GCs while finding better alternatives. Opinions differ on what the research says on toxicity and dosage and whether a long-term, low-dose option is viable. Alternative therapies are being studied, but those endeavors are still in the early stages of development.
While GCs are still used chronically in many patients, clinicians should always attempt to discontinue them whenever possible, Frank Buttgereit, MD, professor of rheumatology and deputy head of the Department of Rheumatology and Clinical Immunology at Charité – Universitätsmedizin Berlin, Berlin, Germany, told attendees of EULAR 2024. Up to 60% of patients in registries use GCs, and many patients with early or established RA enter randomized controlled trials on GCs as maintenance therapy.
The ubiquity of GC usage stems in part from overprescribing by non-rheumatologist physicians who might not have access to or aren’t aware of newer biologics or disease-modifying antirheumatic drugs (DMARDs). “We see a lot of patients on long-term glucocorticoids, chronic use for years and years, decades of glucocorticoids,” said Giovanni Adami, MD, PhD, a rheumatologist at the University of Verona, Verona, Italy, who has coauthored several studies on the use of GCs.
Societies Agree: Discontinue as Fast as Possible
GCs have been associated with a long list of adverse events, most notably Cushing syndrome, hypertension, cardiovascular disease, osteoporosis, myopathy, peptic ulcer, adrenal insufficiency (AI), infections, mood disorders, ophthalmologic disorders such as cataracts, skin disorders, menstrual septic necrosis, and pancreatitis.
Dose matters, Smolen said, citing studies that found that cumulative GC doses of 1000 or 1100 mg increase risks. One study by German researchers found that doses above 10 mg/d significantly raised the hazard ratio for death.
Because high disease activity is also associated with an equally high mortality risk, “we have to balance this out: Active disease vs glucocorticoid use, especially in countries that have less access to modern therapies than we have in the more affluent Western regions,” Smolen said.
Rheumatology societies generally agree that clinicians should try to minimize GC use or eventually discontinue the therapy.
The American College of Rheumatology recommends not using GCs as part of the first-line treatment of RA. “And if you want to use [them], you should do that for less than 3 months, taper and discontinue as fast as possible, and use the lowest dose possible,” Adami said.
EULAR’s recommendation is more nuanced in that it allows for a lower dose but gives physicians more choice in how they want to handle GCs, Adami said. The task force added that all patients should try to taper down or discontinue as fast as possible, he said.
For GCs in the management of systemic lupus erythematosus, a EULAR task force recommended that the type and severity of organ involvement should determine dose, with a long-term goal of maintaining the dose < 5 mg/d or possibly withdrawing it.
EULAR also recommends GC bridging when initiating or changing conventional synthetic (cs) DMARDs. This effectively dismisses the use of GCs when using biologic DMARDs or targeted synthetic DMARDs. As a bridging therapy, EULAR recommends either a single parenteral dose of GC or a predefined tapering or discontinuation scheme within 3 months, when starting an oral GC.
Low-Dose Approach Gains Ground
While saying he’d be the first physician to eliminate GCs whenever possible, Buttgereit made the case before the EULAR Congress that GCs in low doses could still play a role in treatment.
Many physicians believe that very low doses between 2 and 4 mg/d are a realistic therapy option for RA, he said, adding that a mean daily usage < 5 mg could be used over a longer period with relatively low risk.
Several studies he coauthored tested the 5-mg approach. The GLORIA trial compared 5 mg/d prednisolone and placebo in 451 patients aged 65 years and older with active RA over the course of 2 years. The researchers found that patients on prednisolone had a mean Disease Activity Score in 28 joints (DAS28) that was 0.37 points lower and mean joint damage score that was 1.7 points lower than those of patients on placebo, suggesting that the GC had long-term benefits in these patients with RA.
The tradeoff was a 24% increase in the risk of having at least one adverse event of special interest, but most of these events were non-severe infections, Buttgereit said.
Another study, the SEMIRA trial, assigned 128 patients to a continued regimen of prednisone 5 mg/d for 24 weeks. Another group of 131 patients received a tapered-prednisone regimen. All patients received tocilizumab 162 mg with or without csDMARDs, maintained at stable doses.
Patients in the first cohort achieved superior disease activity control than those in the tapered regimen group. “The side effects showed that in the tapering prednisone group, there were more treatment-emergent adverse effects in this double-blind trial as compared to the continued prednisone group,” Buttgereit said.
One limitation of the SEMIRA trial was that it studied the effect of tocilizumab as a GC-sparing agent, and it didn’t consider using a tumor necrosis factor or Janus kinase (JAK) inhibitor, which might have a more potent effect on pain and GC dose reduction, Adami said. “Why do we need to use glucocorticoids if we know they might be detrimental, if we know there might be some other option in our armamentarium?”
Other studies have shown that low-dose GC protocols can be used with standard treatment, according to Sebastian E. Sattui, MD, assistant professor of medicine and director of the Vasculitis Center at the University of Pittsburgh School of Medicine, Pittsburgh.
“Examples of this are the LoVAS and PEXIVAS studies for antineutrophil cytoplasmic antibody-associated [ANCA] vasculitis. This has been highlighted in existing treatment recommendations for ANCA vasculitis and systemic lupus erythematosus nephritis,” Sattui said.
Two-year results from LoVAS showed noninferiority in remission induction rates and rates of relapse and significantly less frequent serious adverse events between a reduced-dose GC regimen at 0.5 mg/kg/d and conventional high-dose GC regimen at 1 mg/kg/d plus rituximab for ANCA vasculitis.
PEXIVAS demonstrated the noninferiority of a reduced-dose regimen of GCs vs a standard-dose regimen with respect to death or end-stage kidney disease in patients with severe disease involvement.
Debating the Toxicity Threshold
Are low GC dosages significantly associated with adverse events like mortality, cardiovascular, or diabetes risk? It depends on who you ask.
Much of the toxicity data on GCs come from inadequately powered or controlled studies and often refer to doses that currently are considered too high, Buttgereit said. His presentation highlighted a study from Hong Kong, a time-varying analysis of GC dose and incident risk for major adverse cardiovascular events (MACE) in more than 12,000 patients with RA. Researchers found that GC regimens ≥ 5 mg/d significantly increased the risk for MACE. Comparatively, doses below this threshold did not confer excessive risk, he said.
Low-dose GCs are lesser toxic than high-dose GCs, noted Joan Merrill, MD, a professor with the Arthritis and Clinical Immunology Research Program at The University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma. “There may be less weight gain, less chance of acne, and less risk for all the slower, more organ-threatening side effects.”
Merrill, who cares for patients with lupus, said physicians can keep lupus in check for years, using constant, low-dose GCs. “The one thing we know is that steroids work.” But over many years, damage may still occur, she cautioned.
But even a low dose could present health problems to patients. The GLORIA trial of patients with RA, which showed promising results on disease control with 5 mg/d, found an association between GCs and increased risk for infection and osteoporosis. There was a higher overall risk for adverse events related to skin, infections, and bone mineral density changes. Bone mineral density loss and fractures were more common in the GC group, Adami noted.
Surprisingly, some of the trial’s authors said patients could handle such adverse events. But what is your threshold of “acceptable?” Adami asked.
Other studies have found associations between low-dose GC regimens and adverse events. Researchers of a 2023 study reported bone mineral density loss in patients with inflammatory rheumatic musculoskeletal diseases on a 2.5-mg/d regimen. Another decade-long analysis of Medicare and Optum data found a link between serious infection and low-dose GCs in patients receiving stable DMARD therapy. Investigators reported risk even at daily doses of ≤ 5 mg.
Adami acknowledged that these studies may have “confounding by indication,” a channeling bias in which people with severe RA are more likely to be treated with GCs. For this reason, it’s a challenge to disentangle the independent role of GCs from the disease activity itself, he said.
The big question is: Why don’t these observational studies show an increased risk for adverse events with biologic drugs that are given to more severe patients? “That confirms the hypothesis that confounding by indication for GCs is minimal, and most of the risk is driven by GCs,” he said.
Tapering Options Across Diseases
Rheumatologists in the field continue to navigate GC-tapering options and treatment combinations that reduce the cumulative use of GCs over time, finding their own solutions based on the conditions they treat.
In his EULAR presentation, Buttgereit suggested that current therapeutic approaches for RA may be too narrow when they don’t consider the possibility of including very low doses of GCs.
For RA, “why shouldn’t we not do a combination of something like methotrexate plus a JAK inhibitor or a biological,” plus a very low dose of GCs < 5 mg/d, he asked.
However, Adami said he generally avoids GCs if RA disease activity is not severe (based on DAS28) and if the patient has a visual analog scale pain score < 7. “Nonetheless, even in patients with more severe disease, I would avoid GCs for more than 3 months. Usually, 1 month of steroids, tapered rapidly and discontinued.”
All patients should receive an appropriate treat-to-target strategy with csDMARDs and biologics if needed, he added.
A patient coming to clinic with difficult-to-treat RA who chronically uses GCs deserves special attention. The priority is bone protection with an anti-osteoporosis medication. “I found that JAK inhibitors, in some cases, help with the discontinuation of steroids, especially in those with residual pain. Therefore, I would think of switching medication,” Adami said.
For polymyalgia rheumatica, most clinicians will likely try to taper GCs around 52 weeks, similar to ACR/EULAR guidelines, according to Robert Spiera, MD, director of the Scleroderma and Vasculitis Program at Hospital for Special Surgery, New York City.
“I usually challenge patients with a more rapid taper, hoping to get them off GCs in 6 or even 4 months in some patients, recognizing that many will flare, and we will have to bump up their GC dose,” Spiera said.
For patients with lupus, GCs remain the most effective treatment, Merrill said. “The toxicities are unacceptable for long-term use. So we try to get in fast when we need them and get out as soon as possible after that, tapering down as fast as the patient can tolerate it.”
Unfortunately, that’s not always as fast as the clinician or patient hopes for, she said.
“New treatments are being developed that may help us avoid the constant use of steroids. However, it would be wonderful to see how these new safer types of steroids work in lupus,” she said.
Minimizing GCs is an important goal that should be considered and aimed for in every single patient, Sattui said. “Risk of GC toxicity should be considered in all patients, assessing [them] for cardiometabolic comorbidities, bone metabolic diseases, risk of infection, among many others.” Sticking to one specific GC-tapering protocol might not be achievable for every patient, however, based on disease characteristics, response, and other factors, he added.
Monitoring for GC toxicity is important and should occur during and after every single clinical visit, he emphasized. Patient education is critical. “Different tools have been developed and employed in clinical trials, both patient- and physician-facing instruments. Implementation to clinical practice of some of these should be the next step in order to achieve a more systematic approach.”
What to Consider for AI Symptoms
Clinicians also need to address AI in patients who are coming off GCs, Sattui said. He advised that symptoms suggestive of AI, including malaise, fatigue, nausea, and muscle and/or joint pain, should guide testing.
Even in the absence of symptoms, clinicians should consider assessing patients who have been on high doses for prolonged periods or obese or older adults who might be at a high risk for AI. “Signs to consider include weight loss, hypotension, or orthostatism,” he said.
Differentiating between AI symptoms and symptoms from the underlying disease can be a challenge. This requires a physical exam and workup, including morning serum cortisol. Collaboration with endocrinology colleagues and other treating providers is important, as well as patient education of symptoms and monitoring for possible adjustments in treating AI and other acute diseases, he said.
Smolen received research grants from AbbVie, AstraZeneca, Galapagos, and Eli Lilly. Adami received speaker fees and/or was a consultant for Galapagos, Theramex, Amgen, Eli Lilly, UCB, Fresenius Kabi, Bristol Myers Squibb, Abiogen, and Pfizer. Buttgereit’s disclosures included AbbVie, AstraZeneca, Grünenthal, Horizon Therapeutics, Mundipharma, Pfizer, and Roche. Merrill had no relevant disclosures. Spiera has been a consultant for Roche-Genentech, GlaxoSmithKline, Sanofi, ChemoCentryx, Novartis, Galderma, Cytori, AstraZeneca, Amgen, and AbbVie and received research grant support from GlaxoSmithKline, Roche-Genentech, AstraZeneca, Bristol Myers Squibb, Kadmon, Boehringer Ingelheim, Cytori, ChemoCentryx, Corbus, Novartis, Amgen, and AbbVie. Sattui reported receiving research support from AstraZeneca and GlaxoSmithKline (clinical trials), receiving consulting fees from Sanofi (funds toward research support), serving on advisory boards for Sanofi and Amgen (funds toward research support), and receiving speaker fees from Fresenius Kabi (funds toward research support).
Jennifer Lubell is a freelance medical writer in the Greater Washington area.
 
Send comments and news tips to [email protected].