Glucocorticoids in the treatment of rheumatoid arthritis

J.W.J. Bijlsma1, J.W.G. Jacobs1, F. Buttgereit2

1
Department of Rheumatology and Clinical ABSTRACT ally spoken the balance between effi-
Immunology, University Medical Center Glucocorticoids (GC) are now being cacy and adverse events is a favourable
Utrecht, The Netherlands; used for over 65 years in the treatment one at and below this dosage.
2
Department of Rheumatology & Clinical
of rheumatoid arthritis (RA). There is In 2007 a Cochrane review was pub-
Immunology, Charité Universitätsmedizin
Berlin, Germany. by now good evidence for their disease lished regarding the disease-modifying
modifying effect, especially in early effect of GC in patients with RA (4).
Johannes W.J. Bijlsma, MD
Johannes W.G. Jacobs, MD RA. When used in a dosage of 7.5–10 In all but one trials in RA patients in
Frank Buttgereit, MD mg most adverse effects can be quite which GC were compared to placebo,
Please address correspondence to: well handled, though monitoring and GC significantly retarded the progres-
Hans Bijlsma, MD, awareness for infections are important. sion of erosions, assessed after one
Department of Rheumatology The CAMERA II study is discussed, as well as after two years duration of
and Clinical Immunology, in which patients with early RA were therapy. Remarkably, in different stud-
University Medical Center Utrecht treated with a tight control scheme of ies it was reported that this retardation
Box 85500; 3508 GA Utrecht, climbing dosages of methotrexate plus of erosion progression persisted, even
The Netherlands.
E-mail: [Link]@[Link]
either 10 mg prednisone daily or place- years after stopping the GC (5, 6); this
bo. After the two years of the trial, 70 % effect has not been shown for any other
Received and accepted on August 28, 2015.
of the patients treated with tight control disease modifying antirheumatic drug
Clin Exp Rheumatol 2015; 33 (Suppl. 92): strategy without GC had no erosions (DMARD). It may be suggested that
S34-S36.
versus 82% of the patients treated with GC have a greater beneficial effect on
© Copyright Clinical and
additional prednisone. Remission was joint structure than can be explained by
Experimental Rheumatology 2015.
reached more often and earlier on in their anti-inflammatory effects only.
Key words: glucocorticoids, the strategy with prednisone compared Adverse events of low dose glucocor-
disease modifying antirheumatic to the strategy with placebo. It may be ticoids are well known, but have to be
drugs, rheumatoid arthritis, suggested that GC have a greater ben- evaluated in the context of the treated
tight control, methotrexate eficial effect on joint structure than can inflammatory condition. Chronic in-
be explained by their anti-inflammatory flammation increases the risk of cardi-
effects only. ovascular diseases, of infections, of in-
sulin-resistance, of inflammatory bone
Glucocorticoids (GC) were used in a loss and other comorbidities. Reducing
rheumatoid arthritis (RA) patient for these risks by reducing chronic inflam-
the first time in 1948, with an impres- mation may in some way counterbal-
sive effect, leading to the Nobel prize ance the negative effects of GC; GC are
for Medicine to Kendall, Reichstein so to say a double edged sword (3). For
and Hench in 1950 (1). These “prelimi- instance, oral glucose tolerance tests
nary results” led to intemperate and showed an increased area under the
unscientific extremes of exaggerated curve of blood glucose as well as C-
praise, bitter denunciation and emo- peptide in patients with RA compared
tion-laden criticisms. Though some to healthy controls, but no difference
of the emotions around the use of GC between RA patients on longstanding
have now been tempered, finding the GC treatment and RA patients naïve for
right balance between advantages and GC treatment was observed (7).
disadvantages still is a matter of debate We are now able to manage many of
(2). GC have now reached 65 years of the reported adverse events of GC, e.g.
age, but they are not allowed to retire preventing bone loss by prophylactic
yet, and probably never will. Informa- treatment with bisphosphonates, calci-
tion from European databases indicates um and vitamin D. The clinically most
that about half of all RA patients even important adverse effect to deal with is
now is using concomitant GC therapy the increased infection risk; especially
for a more prolonged period of time, in the elderly population a dose-relat-
nearly all in dosages below 7.5 mg of ed increase in infections is reported
Competing interests: none declared. prednisone equivalent/day (3). Gener- among patients using GC, with an odds

S-34
 Glucocorticoids in the treatment of RA / J.W.J. Bijlsma et al.

ratio (OR) of 1.72–2.26, depending on this double-blind randomised study, all without GC; the start of the first remis-
actual dosage and duration of GC use patients were treated with the computer sion occurred earlier in the prednisone
(8). This increased risk is strongly in- guided monthly tight control scheme, group (6 vs. 11 months from start of
fluenced by additional risk factors, aiming for remission with increas- the trial). If oral MTX was not well
such as higher age of the patients (>60 ing dosages of MTX, but randomised tolerated, or not leading to remission,
years), chronic lung or renal disease, to addition of 10 mg prednisone daily MTX was given subcutaneously (sc);
low physical function, history of seri- or of placebo from start for the whole if thereafter remission was not reached,
ous infection and concomitant other study duration of 24 months (14). All adalimumab was added as an addition-
medication (9). Also because signs and patients had early RA according to the al step. In the prednisone group only 26
symptoms of infection may be blunted ACR 1987 criteria, with a disease du- patients needed sc MTX versus 60 in
by the use of GC, a high rate of sus- ration less than one year, and all were the placebo group; the difference in use
picion and, if needed, monitoring for DMARD and GC naïve. Treatment of the biological was even more im-
infections in patients on chronic GC was started with MTX 10 mg/week, if pressive: 42 in the strategy group with
treatment are important (10). From a necessary the dose was increased every placebo needed it versus only 16 in the
large observational cohort study in the month with 5 mg up to 30 mg/week; if strategy group with prednisone. The
UK it has become evident that low to no remission was reached at the maxi- more remarkable is the favourable re-
medium dosages of GC are associated mum MTX dose, adalimumab 40 mg/2 sult in erosion score in the latter group.
with an increase in heart failure (OR weeks was added. In addition either 10 A detailed register of adverse events
1.18, confidence interval (CI) 1.05- mg prednisone or placebo was added was kept at each monthly visit; in the
1.33), but NOT with myocardial infarc- from start of the trial. When remission prednisone strategy group, there was
tion, stroke, transient ischemic attack was present for more than 3 months, the no increase in infections, cardiovascu-
or cardiovascular mortality (11). The treatment was reduced, first the MTX. lar events, new diabetes mellitus, new
relative risk for cardiovascular events The definition of remission that was hypertension or fractures. There were
in patients taking high dosages of GC, used in both CAMERA studies was: significantly less gastrointestinal ad-
however, was 2.56 (CI 2.18–2.99). no swollen joints and 2 out of the fol- verse effects in the prednisone strategy
Tight control treatment of patients with lowing 3: tender joints 3 or less, ESR group compared to the placebo strat-
RA, especially early RA, is strongly 20 or less, VAS general health (0-100 egy group, especially nausea (reported
promoted, based on insight into the mm, 100 being the worst score) 20 or at least once during the trial by 23 pa-
disease process (“window of opportu- less. Randomised were 236 patients, of tients versus 43 patients, respectively),
nity”), based on the results of different whom 60 % was woman; mean age was and remarkably less liver enzyme dis-
clinical trials and on consensus meet- 54 years; 68% of patients was rheuma- turbances (ALAT above upper limit of
ings (12). One of the studies that clear- toid factor positive; the mean ESR was normal): assessed at least once during
ly confirmed the superior value of tight 35 mm; patients had a mean of 16 ten- the trial in 15 patients versus 33 pa-
control was the CAMERA (Computer der and 15 swollen joints. Primary out- tients, respectively. GC seem to im-
Assisted Management in Early Rheu- come was the Sharpvander Heijde ero- prove the gastrointestinal tolerance of
matoid Arthritis) study, performed in sion score at the end of the two years; MTX; this could have been due to less
the region of Utrecht, the Netherlands. this primary outcome was statistically NSAID use in the prednisone group
In this study all patients were treated significantly in favour of the GC group. because of lower disease activity (as
with the same drug treatment aiming Importantly, after the two years of the we demonstrated in our earlier Utre-
at remission (step-up methotrexate trial, 70% of the patients treated with cht study, in which we compared the
(MTX) therapy, to which cyclosporine tight control strategy without GC had effects of GC with those of placebo in
was added in case of inefficacy), but no erosions versus 82% of the patients early RA (15)), and also to less use of
they were randomised to one of two treated with additional prednisone. As MTX in the prednisone group although
treatment strategies: tight control expected, clinical variables, as well as the difference was small: mean weekly
(evaluation every month, and in case CRP and ESR improved during the first MTX dosage in the prednisone strategy
of inadequate improvement increase 6 months more in the strategy with GC group 20 mg, versus 23 mg in the place-
in treatment, and in case of remission than in the strategy without GC; after bo strategy group. A third explanation
reduction in treatment; the decision 6 months the variables were similar in could be direct interaction between GC
determined with a computer program) both groups, as a consequence of the and MTX at the cellular level, as has
versus usual care at that time (evalu- continued striving for remission. ACR been suggested in in-vitro studies (16).
ation every three months, increase or 50% response after 1 year and ACR With regard to adverse events on bone,
decrease dependent on the judgement 70% response after two years were sig- we measured bone density with DEXA
of the treating rheumatologist, albeit nificantly better in the prednisone treat- and monitored for (peripheral) frac-
aimed also at remission) (13). ed group. Remission during at least 3 tures (17). We did not find any differ-
The second CAMERA study (CAM- months was reached at least once in ence between the prednisone strategy
ERA II) performed in the same Utrecht 72% of patients in the strategy with group and the placebo strategy group.
region has recently been published. In GC versus 61% of those in the strategy In fact bone mineral density increased

S-35
Glucocorticoids in the treatment of RA / J.W.J. Bijlsma et al.

significantly in both groups, but we ed to occur during treatment with TNF 10. van der GOES MC, JACOBS JW, BOERS M et
al.: Monitoring adverse events of low-dose
treated all patients with daily calcium alpha blockers; this might also be the
glucocorticoid therapy: EULAR recommen-
and vitamin D and the bisphosphonate case for treatment with GC. dations for clinical trials and daily practice.
alendronate 70 mg/week, according to Ann Rheum Dis 2010; 69: 1913-9.
Dutch guidelines (18). References 11. WEI L, McDONNALD TM, WALKER BR: Tak-
ing glucocorticoids by prescription is associ-
Patients on GC often worry about 1. HENCH PS: The reversibility of certain rheu-
ated with subsequent cardiovascular disease.
weight gain. Indeed, in the CAMERA matic and non-rheumatic conditions by the
Ann Intern Med 2004; 141: 764-70.
use of cortisone or of the pituitary adrenocor-
II study patients in the prednisone strat- 12. SMOLEN JS, ALETAHA D, BIJLSMA JW et al.:
ticotropic hormone: Nobel Lecture. In: Les
egy group gained more weight than Prix Nobel en 1950. Stockholm: PA Norstedt
Treating rheumatoid arthritis to target: rec-
ommendations of an international task force.
those in the placebo strategy group & Söner, 1951: 195-223. Ann Rheum Dis 2010; 69: 631-7.
(mean 2.9 kg versus 1.3 kg; p=0.03). 2. BOERS M, KIRWAN JR, BIJLSMA JWJ: Ameri- 13. VERSTAPPEN SM, JACOBS JW, van der VEEN
can College of Rheumatology treatment
Data were analysed with a longitudi- MJ et al.: Intensive treatment with methotrex-
guidelines continue to omit guidance on glu- ate in early rheumatoid arthritis: aiming for
nal regression (mixed model) analysis cocorticoids; comment on the article by Sin- remission. Computer Assisted Management
with BMI as the dependent variable gh et al. Arthritis Care Res 2012; 64: 1622. in Early Rheumatoid Arthritis (CAMERA,
and treatment strategy and DAS28 as 3. BUTTGEREIT F: Do the treatment with glu- an open-label strategy trial). Ann Rheum Dis
cocorticoids and/or the disease itself drive
independent variables, correcting for 2007; 66: 1443-9.
the impairment in glucose metabolism in pa- 14. BAKKER MF, JACOBS JW, WELSING PM et al.:
baseline BMI and possible confounders tients with rheumatoid arthritis? Ann Rheum Low-dose prednisone inclusion in a metho-
(sex, age, and rheumatoid factor status) Dis 2011; 70: 1881-3. trexate based, tight control strategy for early
(19). No independent association was 4. KIRWAN JR, BIJLSMA JWJ, BOERS M, SHEA rheumatoid arthritis: a randomized trial. Ann
BJ: Effects of glucocorticoids on radio-
found of GC therapy with a change in Intern Med 2012; 156: 329-39.
logical progression in rheumatoid arthritis. 15. van EVERDINGEN AA, JACOBS JW, SIEW-
BMI, but a lower DAS28 was associ- Cochrane Database Syst Rev 2007; 241: ERTSZ van REESEMA DR et al.: Low-dose
ated with an increased BMI 6 months CD006356. prednisone therapy for patients with early
later. Clinical cut-off points showed a 5. LANDEWÉ RB, BOERS M, VERHOEVEN AC et active rheumatoid arthritis: clinical efficacy,
al.: COBRA combination therapy in patients
clear association between DAS28 level disease-modifying properties, and side ef-
with early rheumatoid arthritis: long-term fects: a randomized, double-blind, placebo-
and the change in BMI 6 months later. structural benefits of a brief intervention. Ar- controlled clinical trial. Ann Intern Med
Weight gain during treatment with thritis Rheum 2002; 46: 347-56. 2002; 136: 1-12.
prednisone in early RA thus seems at- 6. JACOBS JW, van EVERDINGEN AA, VER- 16. GATICA H, ALISTE M, GUERRERO J et al.:
STAPPEN SM, BIJLSMA JW: Followup ra-
tributable to a reduction of disease ac- Effects of methotrexate on the expression of
diographic data on patients with rheumatoid the translational isoforms of glucocorticoid
tivity and is probably, at least partly, arthritis who participated in a two-year trial receptors α and β: correlation with metho-
regaining weight that was lost when the of prednisone therapy or placebo. Arthritis trexate efficacy in rheumatoid arthritis pa-
RA was not yet adequately treated. Rheum 2006; 54: 1422-8. tients. Rheumatology 2011; 50: 1665-71.
7. HOES JN, van der GOES MC, van RAALTE 17. van der GOES MC, JACOBS JWG, JURGENS
In conclusion, GC have a clear
DH et al.: Effects of chronic low-to-medium MS et al.: Are changes in bone mineral den-
DMARD effect in early RA; a starting dose glucocorticoids on glucose tolerance, sity different between groups of early rheu-
dosage ranging from 7.5 to 10 mg pred- insulin sensitivity and beta-cell function in matoid arthritis treated according to a tight
nisone equivalent seems best regarding chronic rheumatoid arthritis patients. Ann control strategy with or without prednisone if
Rheum Dis 2011; 70: 1887-94. osteoporosis prophylaxis is applied? Osteo-
efficacy as reported in literature and the
8. DIXON WG, KEZOUH A, BERNATSKY S et poros Int 2013; 24: 1429-36.
balance between efficacy and safety. al.: The influence of systemic glucocorticoid 18. HOES JN, JACOBS JW, BOERS M et al.:
Adverse effects tend to be overestimat- therapy upon the risk of non-serious infec- EULAR evidence-based recommendations
ed; the clinically most relevant adverse tion in older patients with rheumatoid arthri- on the management of systemic glucocor-
tis: a nested case-control study. Ann Rheum ticoid therapy in rheumatic diseases. Ann
event is the increase in infection rate,
Dis 2011; 70: 956-60. Rheum Dis 2007; 66: 1560-7.
especially in elderly patients. It may be 9. STRANGFELD A, EVESLAGE M, SCHNEIDER 19. JURGENS MS, JACOBS JW, GEENEN R et
suggested that GC have a greater ben- M et al.: Treatment benefit or survival of the al.: Increase of body mass index in a tight
eficial effect on joint structure than can fittest: what drives the time-dependent de- controlled methotrexate-based strategy with
crease in serious infections rates under TNF prednisone in early rheumatoid arthritis: side
be explained by their anti-inflammatory
inhibition and what does this imply for the effect of the prednisone or better control of
effects only. Uncoupling of disease ac- individual patient? Ann Rheum Dis 2011; 70: disease activity? Arthritis Care Res 2013; 65:
tivity and joint damage has been report- 1914-20. 88-93.

S-36