Rheumatology as a specialty has undergone dramatic transformation in the available armamentarium for treatment options during the past seven decades, ever since the use of corticosteroids was first described in rheumatoid arthritis (RA). However, the recent article by Wang and Panush in the journal suggests that little has changed during this time with regard to corticosteroid use [1]. Particularly concerning was the fact that nearly two-thirds of patients with systemic lupus erythematosus (SLE) admitted to their wards were on high-dose corticosteroids, similar to the proportion of patients in a historical control of lupus patients from the 1950s [1, 2]. The comparisons for patients with RA were probably hindered by the smaller number of patients (nine), most of whom were on corticosteroids, when compared with nearly two-thirds in the 1980s and early 1990s and one-third in the late 1990s and 2000s [1, 3]. However, there were issues associated with the chosen historical controls [4]. Although both lupus cohorts were hospital-based, the clinical characteristics of the lupus cohort in the 1950s as well as in the present study were not available; hence, inferences about severity of disease and the need for immunosuppressive therapy could not be made [1, 2]. The comparisons for RA patients, based on hospitalized patients in the present study with population-based historical cohorts, were more difficult to justify [1, 3]. Nevertheless, the authors have raised an important point about the need to evaluate critically the status of therapeutic strategies for rheumatic diseases that minimize corticosteroid use. This is particularly an ethical issue in high-income countries; costlier, newer therapies are more accessible than in lesser economically developed regions of the world. In this article, we revisit the adverse effects associated with corticosteroid therapy and critically evaluate how far we have actually reached in our search for lesser evils as alternatives to corticosteroids.

Problems associated with corticosteroid use

Corticosteroids are associated with a number of metabolic adverse effects. Bone loss starts to set in within weeks of corticosteroid use and is related to multiple factors, including reduction of calcium absorption, increased calciuria, and alteration of osteoblast-osteoclast balance. Other musculoskeletal adverse effects include osteonecrosis (particularly in situations associated with underlying hypercoagulable states such as lupus with antiphospholipid antibodies) and steroid-induced myopathy. Patients treated with corticosteroids are prone to develop dysglycemia and hypertension and have a greater predisposition to develop atherosclerosis. Altered fat distribution leads to moon facies, buffalo hump, and cutaneous striae. Ocular side effects include increased intraocular pressures and premature posterior subcapsular cataracts. Initial high-dose corticosteroid therapy can occasionally result in steroid psychosis [5]. There is an increased predisposition towards systemic infections, particularly in those treated with pulse bolus corticosteroids used in life-threatening organ involvement in lupus or small vessel vasculitis [6].

In a large multicentric cohort of 1722 lupus patients, most of whom were young (mean age 35 years) with a disease duration of about 1 year at enrollment, and the use of corticosteroids was associated with 1.64 times higher rate of developing damage features [7]. In another cohort of 735 patients with antineutrophil cytoplasmic antibody (ANCA)–associated vasculitis (AAV) enrolled in various clinical trials conducted by the European Vasculitis Study Group (EUVAS), a significant proportion of treatment-related damage was due to hypertension, diabetes mellitus, and osteoporosis, all of which are known adverse effects of corticosteroid use [8]. Thus, the use of corticosteroids comes at a significant cost and potentially little benefit on long-term outcomes despite apparently controlling disease activity, as highlighted by Wang and Panush [1].

Alternatives to corticosteroids in rheumatoid arthritis

Conventionally, patients with RA generally required on low-dose oral corticosteroids along with conventional disease-modifying antirheumatic drugs (DMARDs), with a tapering and stopping of corticosteroid use by 6–9 months once the effect of DMARDs sets in. Over the past two decades, biological DMARDs (bDMARDs) and lately targeted synthetic DMARDs (tsDMARDs) have been increasingly used to help attain remission or low disease activity state as earlier options after initial failure of methotrexate monotherapy [9]. Overall, bDMARDs probably help achieve remission or low disease activity earlier than cDMARDs, although eventually to a similar degree in both instances [10]. Recent evidence suggests that some tsDMARDs like upadacitinib might actually be superior to bDMARDs for controlling disease activity in RA [11]. It is possible that earlier bDMARD and tsDMARD use might obviate the need for high-dose oral corticosteroids, as were previously used in the COBRA trial [12]. Earlier issues regarding the accessibility to bDMARDs in low- and middle-income countries were mostly based on significantly higher costs as well as concerns about safety [13]. Recent literature has proven to be reassuring in this regard. The availability of biosimilar drugs as more cost-effective alternatives has resulted in a wider access to bDMARDs across the world [14]. Literature regarding the persistence of biosimilars in patients switched from innovator bDMARDs to biosimilars is encouraging. A recent report suggested that nearly 85% patients were able to continue biosimilars 1 year after switching [15]. The endemicity of tuberculosis infection has been another concern in low-middle-income countries because of the increased risk of infections, particularly tuberculosis reactivation, with bDMARDs. However, recent data has been reassuring in this regard. In a cohort of 411 patients from India treated with bDMARDs for various rheumatic diseases, baseline screening identified latent tuberculosis infection (LTBI) in 8%. None of these patients underwent tuberculosis reactivation despite bDMARD use after receiving appropriate chemoprophylaxis for tuberculosis. A further 4% patients who had a negative screen for LTBI at baseline went on to develop clinically evident tuberculosis, suggesting the need for continued vigilance during follow-up of these patients [16]. Another study of 209 patients from the same cohort developed serious infections (other than tuberculosis) in < 2% individuals [17]. Another Indian study screened 730 patients with various rheumatic diseases for LTBI using both interferon gamma release assay (IGRA) and Mantoux test (with 10 Tuberculin units) and identified LTBI in 37% individuals, who received appropriate chemoprophylaxis prior to initiating bDMARDs. This cohort of 730 patients was compared with 2930 others treated at the same center without bDMARDs. Both groups had similar proportions of patients developing clinical tuberculosis (about 0.6%) on follow-up [18]. Thus, appropriate screening strategies for detecting LTBI before initiating bDMARDs helped minimize the risk of incident tuberculosis infection (de novo or reactivation) even in high-endemic regions for tuberculosis.

Alternatives to corticosteroids in systemic lupus erythematosus

Numerous efforts have been made to minimize corticosteroid use in lupus due to the aforementioned problems of higher rates of damage accrual in patients on glucocorticoids [7]. The RITUXILUP trial attempted a regimen for remission induction in lupus nephritis with rituximab and mycophenolate mofetil but without oral corticosteroids in 50 patients. The results were highly encouraging, with 90% patients attaining remission by a median of 37 weeks. However, a majority of the cohort comprised membranous lupus nephritis rather than proliferative lupus nephritis (which is more severe), thereby limiting the generalizability of the study findings to more severe lupus nephritis phenotypes [19]. The antagonist of B cell activation factor, belimumab, has demonstrated steroid-sparing effect across multiple lupus trials [20], hence holds promise for evaluation as an alternative to corticosteroid in lupus patients at initial presentation, at least in regions of the world where cost of therapy is not such a significant constraint. There is some evidence that therapies such as tacrolimus targeting p-glycoprotein expression on lymphocytes, which is a marker of steroid resistance, might help reduce corticosteroid dose requirement [21]. Tacrolimus is already proven to be of benefit for the induction of remission in lupus nephritis, particularly in Asian populations [22]. The role of tacrolimus as a cheap steroid-sparing agent in SLE requires wider exploration in settings where access to bDMARDs is limited.

Alternatives to corticosteroids in AAV

ANCA-associated vasculitis is probably the one disease where pauci-steroid regimens have shown the greatest promise. In a pilot study of 20 patients with AAV from North America, a regimen with intravenous rituximab and oral corticosteroids tapered and stopped over 8 weeks could help attain remission in 14/20 patients. The limitation of this study was that patients with more severe AAV manifestations such as severe pulmonary hemorrhage or rapidly progressive renal failure were excluded [23]. Another study from the UK utilized a combined regimen of cyclophosphamide and rituximab for remission induction in AAV. Corticosteroids were only administered for 1–2 weeks in these 49 patients. Of these, 96% were in remission at 6 months and 90% at 12 months [24]. Although some questions have been raised regarding the lower disease severity in this cohort of patients [25], the findings are encouraging regarding the feasibility of reducing steroids in induction regimens of AAV. Greater understanding of the pathogenesis of AAV has revealed alternative complement factor pathway activation as a major event in neutrophil priming in AAV [26]. The complement 5a receptor inhibitor avacopan has been recently tried for remission induction in AAV. An earlier phase II trial suggested the feasibility of a regimen without corticosteroids for remission induction in AAV [27]. The preliminary results of the recently completed phase III trial of avacopan in 330 patients with AAV (166 treated with avacopan, 164 treated with corticosteroids) along with cyclophosphamide (followed by azathioprine) or rituximab identified avacopan as non-inferior to prednisone at 26 weeks for attainment of remission. At 52 weeks, avacopan was superior to prednisone, wherein 66% patients in the avacopan arm attained remission compared with 55% in the prednisone group. Importantly, avacopan continued to have better improvement of renal function than prednisone even in those with severe renal impairment at baseline [28]. These findings hold reasonable promise for a future in the management of small vessel vasculitis management without corticosteroids even in the remission induction regimen.

Future perspectives

The authors would opine that there is considerable promise for corticosteroid-naïve or pauci-corticosteroid regimens for the management of common rheumatic diseases like RA as well as serious multisystemic rheumatic diseases like lupus and AAV. However, there remains a need to evaluate optimal regimens for such situations, which are likely to be disease-specific. Until such regimens are widely accepted, every attempt should be made by treating rheumatologists to minimize corticosteroid dose and duration, as well as institute appropriate protective strategies for bone health while monitoring for metabolic complications, vascular health, and other adverse effects of corticosteroids.