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


Rheumatoid arthritis and sulfasalazine toxicity: a meta-analysis

ACP J Club. 1993 Sept-Oct;119:47. doi:10.7326/ACPJC-1993-119-2-047

Source Citation

Wijnands MJ, Van't Hof MA, Van De Putte LB, Van Riel PL. Rheumatoid arthritis: a risk factor for sulphasalazine toxicity? A meta-analysis. Br J Rheumatol. Apr



To compare patients with rheumatoid arthritis (RA) with patients with inflammatory bowel disease (IBD) and seronegative spondylarthropathies (SSpA) for the occurrence of adverse effects during sulfasalazine treatment.

Data sources

A MEDLINE search from 1980 forward retrieved articles in English, French, and German using the terms clinical trial, rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis, inflammatory bowel disease, and salicylazosulfapyridine. Personal files were also searched.

Study selection

Studies were selected if they had 10 patients, the patients and sulfasalazine therapy were adequately described, all patients were accounted for, and all clinically relevant outcomes were reported. Abstracts, letters, and duplicate publications were excluded.

Data extraction

Number of patients treated, study quality, types of adverse effects, and number of people who discontinued sulfasalazine therapy because of adverse effects.

Main results

49 studies were included: 9 on SSpA with 184 patient-years of study, 17 on RA with 552 patient-years, and 23 on IBD (18 on ulcerative colitis, 4 on Crohn disease, and 1 on both) with 713 patient-years. For side effects that led to termination of treatment, patients in the 3 disease groups did not differ for number with gastrointestinal effects, cutaneous reactions, or headaches. Patients with IBD did not have any hematologic or hepatic adverse effects causing discontinuation of treatment. The relative risk (RR) for discontinuing sulfasalazine treatment because of adverse effects for patients with RA compared with IBD was 4.4 (95% CI 2.8 to 6.8; P < 0.0001) and for RA compared with SSpA was 1.6 (CI 1.07 to 2.24; P < 0.05). These differences disappeared (after adjusting for patient age, previous treatment with sulfasalazine, and dose), except for the treatment termination because of adverse effects for RA compared with SSpA for the first 3 months (RR 0.12, CI 0.05 to 0.29; P < 0.01). Sensitivity analysis showed that study quality did not affect RR or significance calculations.


The occurrence of adverse effects did not differ for patients who had rheumatoid arthritis, inflammatory bowel disease, or seronegative spondylarthropathies (after adjustment for age, previous treatment, and dose), except that patients with seronegative spondylarthropathies discontinued treatment more often than did patients with rheumatoid arthritis in the first 3 months of treatment.

Source of funding: The Dutch League Against Rheumatism.

For article reprint: Dr. M.J. Wijnands, University Hospital Nijmegen, P.O. Box 9101, 6500 HB Nijmegen, Netherlands. FAX 31-80-54-1433.


Sulfasalazine was originally developed to treat RA but became popular in the treatment of IBD. Although S-aminosalicylic acid is the active moiety in IBD, sulfapyridine is the active component in RA. Sulfasalazine is effective for RA and for SSpA, including ankylosing spondylitis and the Reiter syndrome.

This meta-analysis by Wijnands and colleagues summarizes the incidence of severe side effects that caused patients to stop treatment. The most common side effects leading to dropout were gastrointestinal (approximately 46% of treated patients) and cutaneous (approximately 27%). The authors found the side-effect rates of sulfasalazine treatment did not differ substantially in patients with IBD, RA, or SSpA.

Because clinical trials were evaluated, patients were generally treated for a short time (on average, two thirds of a year), and the rules of the trials allowing discontinuation of therapy may have differed from clinical practice and may have differed from one disease to another.

The authors also searched for other factors that affected side-effect rates and found that older patients had a higher risk for toxicity, as did those taking sulfasalazine for the first time and those receiving relatively high doses.

Other factors possibly associated with sulfasalazine toxicity, and not easily evaluable in this study, include whether the drug was enteric coated (most sulfasalazine in the United States is coated) and whether the patients concurrently used other medications likely to induce gastrointestinal side effects, such as nonsteroidal anti-inflammatory drugs.

David Felson, MD, MPH
Boston University Boston, Massachusetts, USA