Current issues of ACP Journal Club are published in Annals of Internal Medicine


Prednisolone reduced the progression of early rheumatoid arthritis

ACP J Club. 1996 Jan-Feb;124:5. doi:10.7326/ACPJC-1996-124-1-005

Source Citation

Kirwan JR and the Arthritis and Rheumatism Council Low-Dose Glucocorticoid Study Group. The effect of glucocorticoids on joint destruction in rheumatoid arthritis. Engl J Med. 1995 Jul 20;333:142-6. [PubMed ID: 7791815]



To determine the effectiveness of a single daily dose of prednisolone in patients with rheumatoid arthritis (RA).


2-year randomized, double-blind, placebo-controlled trial.


13 centers in the United Kingdom.


138 patients (mean age 49 y, 64% women) who were between 18 and 69 years of age and who had currently active RA that developed < 2 years previously. Follow-up was 83%.


Patients were allocated to prednisolone, 7.5 mg/d (n = 61), or to placebo (n = 67). The dose was tapered after 2 years to every other day for 2 weeks, then to every third day for 2 weeks. Patients could receive any medications except systemic corticosteroids.

Main Outcome Measures

Disease progression and development of erosions determined by radiography of the hands (using the Larsen Index). A radiologist and a rheumatologist who were both unaware of patient information jointly read radiographs in randomly ordered groups of 30. Larsen units were used to grade joint damage (0 = normal joint; 5 = maximal destruction), and the Index (ranging from 0 to 140) was calculated for each radiograph. The condition of each hand was classified as erosive or nonerosive.

Main Results

Patients who received prednisolone had an increase from baseline of 0.73 Larsen units at 1 year and an increase of 0.72 units at 2 years; patients who received placebo had increases of 3.63 and 5.73 Larsen units, respectively. When the Larsen scores were log-transformed for between-group comparisons, the mean change at 1 year for patients who received prednisolone was 0.02 (SD 0.39); for placebo recipients, the mean change was 0.19 (SD 0.48) (95% CI for the 0.17 difference 0 to 0.34, P = 0.052). At 2 years, the mean change was 0.02 (SD 0.43) for patients who received prednisolone and 0.30 (SD 0.52) for patients who received placebo (CI for the 0.28 difference 0.09 to 0.47, P = 0.004). Daily treatment with prednisolone reduced the risk for developing erosions during the ensuing 2 years compared with placebo (P = 0.003) (Table).


In patients with early active rheumatoid arthritis, a single daily dose of prednisolone slowed the progression of the disease and the development of erosions.

Source of funding: Arthritis and Rheumatism Council.

For article reprint: Dr. J. Kirwan, University Department of Medicine, Bristol Royal Infirmary, Bristol BS2 8HW, England, UK. FAX 44-117-928-3841.

Table. Prednisolone vs placebo for early rheumatoid arthritis*

Outcome at 2 y Prednisolone Placebo RRR (95% CI) NNT (CI)
Development of erosions 22% 46% 60% (21 to 71) 4 (3 to 12)

*Abbreviations defined in Glossary; RRR, NNT, and CI calculated from data in article.


Cyclosporine plus methotrexate was superior to methotrexate alone for rheumatoid arthritis

RA is a common disease that can be crippling to patients and frustrating to physicians, who continue to seek effective treatments to change its course. Toward this end, these 2 important papers by Tugwell and Kirwan and their colleagues address the therapy for this illness.

The study by Tugwell and colleagues shows the efficacy of cyclosporine plus methotrexate (vs methotrexate alone) when used in patients with severe RA who partially responded to only methotrexate. Nonsteroidal anti-inflammatory drugs are the first-choice drug for early RA. When these agents are insufficient, rheumatologists generally move on to monotherapy with methotrexate, antimalarial agents, gold, penicillamine, sulfasalazine, or more recently, cyclosporine. Using the model of experts in oncology and infectious diseases, this study proposes a combination of therapies and shows that this approach is better than monotherapy with methotrexate alone. Although elegantly designed, it is important to recognize what this study does and does not show.

The study by Tugwell and colleagues does not imply that this combination of therapy should be used by primary care physicians to treat patients with uncomplicated RA. First, patients were followed for only 6 months, and the long-term efficacy and toxicity of this combination have yet to be determined. Second, because this is not a benign combination of drugs, this therapy should be used only by physicians who are experienced in its use. Generally, this would mean that this therapy should be administered by rheumatologists only. Third, the patients included in this study had severe RA; therefore, it is unclear whether the results are generalizable to patients with mild disease. Finally, even the study patients who had severe RA should be regarded as a subset because they had already partially responded to methotrexate. This was an appropriate decision given the purpose of the study (i.e., to determine the effectiveness of this combination therapy). It still remains to be determined whether the combined intervention would be effective in patients who have not taken methotrexate or in those whose disease did not respond to methotrexate but who may have less treatable disease. These questions need to be answered before this becomes a standard treatment, even for those with severe RA.

In contrast to the study by Tugwell and colleagues, Kirwan and colleagues studied a milder, more common form of the disease. This study was also well done, although of the 128 patients randomly assigned, 22 to 25 were not available for scheduled radiologic evaluations. Although a clear radiologic response with a slowed progression of RA was observed when prednisolone was given, the clinical importance of this observation is unclear. The authors observed a short-term clinical improvement in their treatment group, but improvement of symptoms did not persist and long-term therapy with corticosteroids cannot be considered totally benign. Thus, although the use of corticosteroids may be physiologically useful, the long-term benefit is not yet established and the true utility of this widely used intervention must still be determined.

Brian L. Strom, MD, MPH
University of Pennsylvania Medical CenterPhiladelphia, Pennsylvania, USA