Low-dose prednisone is widely used for the treatment of rheumatoid arthritis (RA). When used in conjunction with disease-modifying antirheumatic drugs (DMARDs), it has been shown to retard radiologic damage in RA. However, few data are available on the efficacy of prednisone alone in this regard, or its safety. To address this question, van Everdingen, et al (Ann Intern Med 2002; 136:1-12) conducted a two year trial comparing low dose prednisone alone versus placebo for the treatment of early RA.
In a two year, randomized, double-blinded study the efficacy of 10 mg prednisone vs. placebo was examined in 81 subjects. All subjects had been diagnosed with rheumatoid arthritis within the past year and had active disease (>3 tender joints, >3 swollen joints, ESR > 28 mm/hr, > 30 minutes morning stiffness). No subjects had been exposed any DMARD prior to the study.
Patients were randomized to receive either placebo or prednisone 10 mg every morning with 500 mg supplemental calcium. Acetaminophen and NSAID use was allowed as were intraarticular steroid injections. Sulfasalazine was allowed as a rescue medication after six months if the investigator deemed the disease active.
Thirty-five of 41 placebo subjects and 36 of 40 prednisone subjects completed the study. No subject dropped due to lack of efficacy or side effects, despite restrictions on DMARD therapy within the first six months. At baseline, there were no statistically significant differences in the demographic characteristics or disease activity parameters between the two groups, although there were some trends towards greater severity in the placebo group. After 6 months, 39 of 71 completers (20 placebo, 19 prednisone) received sulfasalazine.
At 12 and 24 months, there was statistically significant less radiographic progression in the prednisone group compared to the control group. Prednisone patients progressed at a rate of approximately 8 radiographic (Sharp) units per year, while placebo patients progressed at approximately 15 units per year. Overall, improvements in clinical efficacy parameters were not significantly different between the groups except for grip strength and 28-joint tender score (better in the prednisone group). However, prednisone treated patients took significantly less daily NSAID and acetaminophen, and had significantly fewer steroid articular injections, than the placebo treated subjects. This may have obscured any potential differences in the clinical parameters between the prednisone and placebo groups. Patients receiving prednisone experienced weight gain (77 to 80 kg, p=0.001) and an increase in mean serum glucose levels (5.1 to 5.9 mmol/L, p=0.01), while placebo patients did not. Prednisone treated patients also had an increased number of new vertebral bone fractures than placebo treated patients (5 versus 2).
Low-dose prednisone provides relief from rheumatoid arthritis symptoms particularly within the first six months and substantially inhibits progression of erosive joint damage in patients nave to other DMARD therapies. However, as the authors note, because of its limited disease modifying properties, prednisone should be prescribed in conjunction with other DMARDs.
There are several important notations to make about this study. As the authors duly note, a placebo-controlled study design would not likely be approved now by human subjects or ethics committees. But, when it was initiated in the early 1990s, it was state-of-the-art and clearly yielded valuable information. Without the confounding addition of DMARDs, this study clearly demonstrates the disease modifying potential of low dose steroids. However, subjects receiving 10 mg prednisone daily still progressed at 8 Sharp units/yr. While this was lower than the placebo rate of 15 units/yr, it is much higher than the rates observed in recent studies in patients treated with methotrexate, etanercept, infliximab + methotrexate, or leflunomide. With these DMARDs, rates of progression are in the range of 0-2 units/yr. Each Sharp unit represents the equivalent of one erosion (or worsening of one erosion) or progression in joint space narrowing in one joint. Eight erosions, or progression of narrowing of eight joints, per year as demonstrated on prednisone treated subjects in this study, would not be acceptable given the higher superiority of the DMARDs listed above. Thus, the recommendation by the authors that prednisone should be adjunctive, rather than monotherapy, is appropriate.
The efficacy of prednisone must also be balanced against its potential long-term side effects. In this study, there was a tendency towards hyperglycemia and weight gain. Insofar as active RA can cause weight loss, some of the weight gain in prednisone patients could conceivably have been due to better control of disease activity and return to pre-RA weight rather than purely a side effect of prednisone. The increase in vertebral fractures is particularly concerning, however, and consistent with a recent report that even nasally administered corticosteroids are associated with decreases in bone density. (see report)
Finally, it would have been interesting to see how the addition of sulfasalazine altered the rates of progression in the two groups. Insofar as equivalent numbers of both treatment groups were started on sulfasalazine, its effect is probably equally distributed among the two groups. However, the subset in both treatment groups who received sulfasalazine might be expected to exhibit a slower rate of progression from 6-12 months than they had in the period of 0-6 months.