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


Increased prednisone reduced minor relapses in systemic lupus erythematosus

ACP J Club. 1996 Jan-Feb;124:6. doi:10.7326/ACPJC-1996-124-1-006

Source Citation

Bootsma H, Spronk P, Derksen R, et al. Prevention of relapses in systemic lupus erythematosus. Lancet. 1995 Jun 24;345:1595-9.



To determine whether additional prednisone after an increase in antibodies to double-stranded deoxyribonucleic acid (anti-DNA) reduces relapses in patients with systemic lupus erythematosus (SLE).


Randomized controlled trial (mean follow-up > 556 d).


3 university hospitals in the Netherlands.


46 patients (mean age 36 y, 87% women) who showed an increase in anti-DNA antibodies were selected from a total of 156 patients being followed for SLE. Inclusion criteria were meeting ≥ 4 American College of Rheumatology criteria for SLE, age ≥ 18 years, no pregnancy, and no contraindications to corticosteroids.


Patients received standard care and were seen at least every 3 months. If anti-DNA increased by 25%, patients were assigned to conventional care (n = 24) or early treatment (n = 22). Conventional care included adjusted prednisone and, for clinical relapse, cytotoxic drugs. In the early treatment group, prednisone was increased to 30 mg above baseline (maximum 60 mg/d), tapered by 10 mg every 3 weeks until 30 mg/d was reached, and then tapered by 5 mg every 3 weeks to baseline levels. 85% of patients completed ≥ 6 months of study.

Main Outcome Measures

Major relapses (severe renal or central nervous system disease; severe serositis or vasculitis; hematologic, eye, or lung disease; or heart or muscle disorders) or minor relapses (increase in SLE disease activity scores [SLEDAI] ≥ 2 points within 6 mo).

Main Results

20 of 24 patients (83%) receiving conventional care had relapses (7 major [29%] and 13 minor [59%]), and 2 of 22 patients receiving early treatment had a major relapse (9%) {95% CI for the 74% absolute risk reduction 49% to 88%; relative risk reduction 89%, CI 66% to 97%, P < 0.001; number needed to treat 1, CI 1 to 2}*. Patients receiving early treatment had fewer minor relapses (0% vs 59%, P < 0.001) and took a higher median dose of prednisone (15.3 vs 10.0 mg/d, P = 0.025). The groups did not differ for major relapse (P = 0.12), adverse effects (71% for usual care vs 82% for early treatment), or total cumulative dose of prednisone (4515 vs 8025 mg, P = 0.068).


In patients with systemic lupus erythematosus, treatment with prednisone immediately after an increase in antibodies to double-stranded deoxyribonucleic acid prevented minor relapses without increasing the total cumulative prednisone dose.

Source of funding: Dutch League against Rheumatism.

For article reprint: Dr. H. Bootsma, Department of Internal Medicine, Division of Clinical Immunology, University Hospital, Groningen, Post Office Box 30.001, 9700 RB Groningen, the Netherlands. FAX 31-50-697115.

*Numbers calculated from data in article.


In the 1970s, standardized serologic tests for SLE suggested the possibility of flare prediction and prevention, but attempts to normalize complement and anti-DNA antibody led to severe corticosteroid toxicity and many patients with persistent serologic abnormalities remained clinically well. Others had non-DNA antibody systems, such as anti-Sm antibody. Thus, enthusiasm for serologic flare prediction and prophylaxis waned. New outcome measurements provided a reproducible way to quantitate SLE disease activity. Bootsma and colleagues use modern methods to re-examine the use of flare prediction and prophylaxis. They conclude that prednisone, 30 mg/d, is effective and carries little short-term toxicity in treating asymptomatic increases in anti-DNA antibody.

The data do not tell the entire story. Persons who had previous adverse experience with corticosteroids were excluded. Another 17% who did not have anti-DNA antibody were also excluded. Of 15 major relapses, 2 occurred despite treatment and 6 occurred without being predicted. Thus, half of the major flares were not preventable. The authors state that no toxicity existed, but as many treated patients withdrew from the study as were protected from major flare. Although the cumulative dose of prednisone did not differ between the 2 groups, the mean dose in the treated group was nearly twice that of the untreated, the P value just missing significance (P = 0.068). Finally, because prevention of minor flares is a justification for prophylactic treatment, what was prevented? According to the study protocol, polyarthritis, fatigue, myalgia, and fever were the main symptoms of minor flares. These symptoms can be treated with low-dose prednisone (≤ 10 mg/d). Patients may have been overtreated to prevent minor flares.

Michael D. Lockshin, MD
National Institute of Arthritis and Musculoskeletal and Skin DiseasesBethesda, Maryland, USA