Glucocorticoids in RA

When I teach trainees about the treatment of rheumatoid arthritis (RA), I always start with questions about glucocorticoids, asking about the first therapeutic use of these agents. Addison’s disease and Cushing syndrome are the usual answers I receive, but I tell the trainees it was another disease.

I then give a hint, “Clinical observations on patients with RA provided a decisive clue,” and I ask, “What could happen in RA to suggest a role of hormones?” If I don’t receive an immediate response, I ask, “Who gets RA most often, men or women? What can happen to a woman to improve RA?”—sometimes adding, “something that can’t happen to a man?”

Eventually, the trainee realizes I am referring to pregnancy. Finished with the Socratic method, I then describe research at the Mayo Clinic on compound E (i.e., cortisone) to treat RA based on the prescient observation that women with RA can have remissions during pregnancy.¹ After noting the Nobel Prize presented to Kendall, Reichstein and Hench for the discovery of glucocorticoids, I then discuss how this group of drugs is now used by essentially every medical specialty in an extraordinary variety of formulations (e.g., oral, parenteral, topical, inhalational, etc.) and various indications. Further, I explain that glucocorticoids are given in doses ranging over three orders of magnitude. For rheumatologists, these doses extend from 1 to 2 mgs of prednisone a day for chronic RA—a “whiff” of steroids as we used to say—to a 1 g pulse of methylpredisolone (Solu-Medrol) for rapidly progressive glomerulonephritis in systemic lupus or vasculitis.

The meaning of low-, moderate- and high-dose glucocorticoids remains vague and should probably include categories for very low and very high. Nevertheless, I find this dose range remarkable because it far exceeds that of any other drug. I often wonder what happens at 1 g of methylprednisolone that does not happen at 1 mg of prednisone. Interestingly, a study of patients treated with tocilizumab indicated that, for some patients, a few milligrams of prednisone provides better disease control than a complete taper, even though blocking interleukin (IL) 6 signaling would seem sufficient to control inflammation.² What does a whiff of prednisone do that a potent biologic can’t?

When I began my career as a rheumatologist, the array of disease-modifying anti-rheumatic drugs (DMARDs) was limited and even those in common use were viewed with skepticism. Although the efficacy of gold salts was established in the Empire Rheumatism Council Trial, many physicians simply did not believe these agents worked.³ Certainly, when methotrexate came along, this group of SAARDs (slow- acting anti-rheumatic drugs, so named to contrast with the more rapid-onset methotrexate) was rapidly abandoned. The treatment pyramid either disappeared or was turned upside down and sideways into another geometry.

In the pre-methotrexate or pre-biologic era, glucocorticoids were often the only way to prevent a patient from going up in smoke— another term from times past. Unlike the SAARD experience, clinical experience with prednisone inspired confidence that inflammation could be quelled. Further, studies showed that glucocorticoids have DMARD action and could attenuate progression of erosion;⁴ nevertheless, glucocorticoids are not usually categorized as DMARDs. For persistently painful joints, intra- articular steroids were often the only way short of surgery (including synovectomy, another relic) to reduce symptoms. Indeed, some patients had multiple injections to provide some comfort over time.

Toxicity

As is now well recognized, gluco- corticoids have a significant short- term and long-term toxicity that can be assessed quantitatively by a glucocorticoid toxicity index.⁵ Without careful dosing (less than 5 to 7.5 mg of prednisone daily), the toxicity can be substantial, especially if higher doses are used for too long a time. Chronic use nevertheless remains a reality, even if guidelines recommend only short-term use for flares or as bridge therapy when starting DMARDs.⁶

The concern about glucocorticoid toxicity has led to an interesting term: steroid-sparing agents. As the name indicates, the desired pharmacologic actions of these agents is to allow glucocorticoids to be reduced or discontinued. I would be reluctant, however, to define effective agents in terms of their activity on other agents. I would not refer to tumor necrosis factor (TNF) inhibitors as methotrexate-sparing agents any more than I would call methotrexate a TNF-inhibitor sparing agent.

Among many older rheumatologists, I find less concern over toxicity than among younger rheumatologists who don’t have to rely on prednisone for treatment benefits. Interestingly, although I prescribe prednisone to be taken in the morning, a colleague (older like me) splits the dose between morning and night, which he claims is more effective. I do wince a bit because I was taught to give prednisone in the morning to mimic the diurnal variation. Chronotherapy is, nevertheless, a concept worth exploring if the usual morning steroid dose is too late to abate the morning burst of cytokines.⁷

The differing perceptions about glucocorticoid safety recently became clear to me when I discussed a patient with one of our fellows. The patient was a man in his 50s who had seropositive RA and persistent disease activity despite 25 mg of methotrexate and 50 mg of etanercept each week. With an elevated Disease Activity Score-28 (DAS28) and a plethora of tender and swollen joints, another agent was clearly indicated.

I asked the fellow his choice, and he surprised me by suggesting a switch from etanercept to rituximab. Although rituximab is an effective agent, I have a general reluctance to kill healthy cells to treat a benign, albeit serious, disease. I especially worry about eliminating circulating B cells for months and the effects on adaptive immune responses, including those to vaccines.

I then asked the fellow if he believes rituximab is safer than 5 mg prednisone and the response was a qualified yes. The fellow then enumerated the usual litany of steroid problems: hypertension, hyperglycemia, osteoporosis, depression, mania and so on. Cardiovascular disease also received mention.

I tried to assure the fellow that, with a low-dose glucocorticoid (i.e., 5 mg of prednisone or prednisolone), those dreaded side effects should be very limited, as shown by studies like the GLORIA trial.⁸ Nevertheless, the teaching against glucocorticoids is so strong that the fellow seemed unconvinced.

One article I read stated the medical community has been “inept” in developing an evidence base on glucocorticoid therapy.⁹ While inept is a strong word, the gaps in knowledge about glucocorticoids remain surprisingly large after 70 years.

In Sum

Because glucocorticoids are going to be part of RA treatment for the foreseeable future, rheumatologists should continue to ask salient questions about their risks and benefits. Hopefully, the answers will be substantive and provide evidence to satisfy even the most demanding attending-cum-Socrates in the clinic.