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


Early SAARDs were better than NSAIDs in recently diagnosed rheumatoid arthritis

ACP J Club. 1996 Sept-Oct;125:36. doi:10.7326/ACPJC-1996-125-2-036

Source Citation

van der Heide A, Jacobs JW, Bijlsma JW, et al. The effectiveness of early treatment with "second-line" antirheumatic drugs. A randomized, controlled trial. Ann Intern Med. 1996 Apr 15; 124:699-707.



To compare the early introduction of slow-acting antirheumatic drugs (SAARDs) with nonsteroidal anti-inflammatory drugs (NSAIDs) alone for disease activity and progression of radiologic abnormalities in recently diagnosed rheumatoid arthritis (RA).


Randomized controlled trial with minimum follow-up of 12 months.


6 rheumatology clinics in the Netherlands.


238 adults (mean age 57 y, 69% women) with recently diagnosed RA (duration < 1 y). Exclusion criteria were age < 17 years; conditions that could interfere with 1 of the treatments; previous or current use of SAARDs, glucocorticoids, or cytotoxic or immunosuppressive therapy; possibility of pregnancy or lactation; or inability to comply. Follow-up was 92%.


57 patients were allocated to the non-SAARD group and received NSAIDs (dose and type could be changed at any time). The SAARDs group (181 patients) had 3 treatment arms: hydroxychloriquine (400 mg/d); intramuscular gold (aurothioglucose, 50 mg/wk); or oral methotrexate (7.5 to 15 mg/wk). If adverse effects occurred, patients could only change SAARDs once (hydroxchloroquine to auranofin 6 to 9 mg/d; intramuscular gold to D-penicillamine, 500 to 750 mg/d; methotrexate to sulfasalazine 2000 to 3000 mg/d). Therapy was continued for 12 months unless adverse effects dictated stopping.

Main outcome measures

Functional disability, pain, joint score, erythrocyte sedimentation rate, and radiologic abnormalities. Secondary outcomes were grip strength, duration of morning stiffness, and general well-being. Hemoglobin and C-protein levels and platelet counts were also monitored.

Main results

Analysis was by intention to treat and the SAARDs groups were analyzed as 1 group. More patients in the non-SAARD group had a defined discontinuation of therapeutic strategy (29% vs 8%, P < 0.05). At 12 months, patients in the non-SAARD group had worse outcomes for disability scores, pain scores, joint score, erythrocyte sedimentation rates, and the secondary outcomes of grip strength, C-reactive protein levels, and platelet counts than patients in the SAARDs group.The groups did not differ for radiologic damage scores, well-being, or duration of morning stiffness.


Early introduction of slow-acting antirheumatic drugs showed more benefit in clinical outcomes than nonsteroidal anti-inflammatory drugs in adults with recently diagnosed rheumatoid arthritis.

Source of funding: Dutch League Against Rheumatism.

For article reprint: Dr. J.W. Jacobs, Department of Rheumatology F02.223, University Hospital Utrecht, PO Box 85500, 3508 GA Utrecht, the Netherlands. Fax 31-10-463-3726.


By examining the effectiveness of early second-line treatment in RA, van der Heide and colleagues address an extremely important concept in the management of rheumatic disease. The strengths of this study include its satisfactory follow-up, intention-to-treat analysis, documentation of patients who refused, and high completion rates. In addition, the guidelines for response and withdrawal are clearly documented. Although reactive arthritis and articular-onset systemic lupus erythematosis are occasionally confused with early-onset RA, only 2 patients were misdiagnosed.

The study also has some evident weaknesses. Using the date of diagnosis of RA rather than the date of onset may be misleading. Although the age and clinical features were consistent with early RA, the proportion of patients who were seropositive was not. 70% to 80% would have been expected but seropositivity was only 60%.

Patients in the non-SAARD group who started taking a SAARD (29%) or increased their steroids reflect the practical realities of treating RA. These medication changes probably "contaminated" the results and made them more conservative. Radiographic changes were not detected because of the small sample size and the unavailability of radiographs for all patients. These problems have also been reported by others. In addition, stratification of patients was problematic for this and other studies (1). Whether the results could be sustained beyond 12 months needs to be shown. It is frustrating that, after patients were randomized to hydroxychloroquine, gold, and methotrexate, results were not presented by individual drug.

The results of the study by van der Heide and colleagues confirm previous reports showing the benefits of SAARDs in early RA, but they also highlight that, for variousreasons, many patients do not agree to a strategy that includes a SAARD. In addition, because early diagnosis of RA can be difficult, universal early treatment with a SAARD may not occur regardless of how advantageous the therapy.

Hilary A. Capell, MD
Royal InfirmaryGlasgow, Scotland, UK


1. Kirwan J. The effects of glucocorticoids on joint destruction in rheumatoid arthritis. The Arthritis and Rheumatism Council Low-Dose Glucocorticoid Study Group. N Engl J Med. 1995;333:142-6.