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


Review: Azathioprine maintains remission in Crohn disease, but withdrawal rates are higher

ACP J Club. 1998 Mar-April; 128:33. doi:10.7326/ACPJC-1998-128-2-033

Source Citation

Pearson DC, May GR, Fick G, Sutherland LR. Azathioprine for maintenance of remission of Crohn's disease. In: The Cochrane Database of Systematic Reviews. The Cochrane Library. Oxford: Update Software; 1997, Issue 3.



To determine, using meta-analysis, whether azathioprine (AZA) maintains remission or facilitates a reduction in steroid dose in adults with Crohn disease.

Data sources

Studies were identified with MEDLINE (1966 to May 1996) by using the terms antimetabolites, azathioprine, mercaptopurine, Crohn disease, and inflammatory bowel disease. Review articles, conference proceedings, and bibliographies of relevant studies were also reviewed.

Study selection

English- and French-language studies were selected if patients were > 18 years of age and had Crohn disease that was in remission (mild or no symptoms regardless of use of prophylactic medications, including prednisone). Studies were randomized, double-blind, placebo-controlled trials that evaluated oral AZA (1.0 to 2.5 mg/kg of body weight per d) for treatment periods of 26 to 52 weeks. Remission was defined by several clinical criteria.

Data extraction

Data were extracted in triplicate on patient and disease characteristics; dose and duration of therapy; proportion of patients who remained in remission, could taper steroid dose, or withdrew from the study because of toxicity; and clinical scores.

Main results

5 trials were included with 319 patients, of whom 130 received AZA. More patients who received AZA remained in remission (P < 0.001) or could taper the steroid dose to < 10 mg/d (P = 0.011) than patients who received placebo. However, more patients who received AZA withdrew from therapy because of adverse events (P = 0.035) (Table).


Patients with Crohn disease in remission who receive azathioprine remain in remission and can reduce the steroid dose, but they have increased rates of withdrawal because of adverse events.

Sources of funding: University of Calgary and Glaxo Canada.

For article reprint: Dr. L.R. Sutherland, Professor of Medicine, University of Calgary, Community Health Sciences, 3330 Hospital Drive North West, Calgary, Alberta T2N 4N1, Canada. FAX 403-270-7307.

Table. Azathioprine vs placebo for Crohn disease*

Outcomes Azathioprine weighted EER Placebo weighted CER RBI (95% CI) Weighted ABI NNT (CI)
Remission 71% 53% 35% (12 to 63) 19% 6 (4 to 12)
Ability to taper steroid dose 83% 53% 63% (1 to 166) 30% 4 (2 to 48)
RRI (CI) Weighted ARI NNH (CI)
Withdrawals for adverse events 6.2% 1.6% 219% (30 to 680) 4.6% 22 (12 to 293)

*Abbreviations defined in Glossary; RBI, ABI, NNT, RRI, ARI, NNH, and CI calculated from data in article.


Pearson and colleagues' methodologically rigorous meta-analysis provides an excellent overview of a topic that has been controversial for almost 2 decades (1, 2). The message for practicing clinicians is that AZA is a safe and effective drug for long-term management of selected patients with Crohn disease when used at an adequate dosage (2.5 mg/kg per day). However, readers are advised to use caution when interpreting the number needed to treat (6 patients need to be treated for 1 additional patient to have a remission) identified by the authors. The clinical trials that were analyzed used widely different measures of clinical response and were heterogeneous with regard to the use of concomitant steroid therapy. Only 2 of the 5 trials analyzed used a generally accepted definition of clinical remission (absence of symptoms [defined by a Crohn Disease Activity Index score < 150] and complete discontinuation of glucocorticoid therapy). Thus, although this report provides additional support for the efficacy of AZA, it is difficult to translate the summary measure of response that was used into clinically meaningful terms. Nevertheless, clinicians who treat patients with severe Crohn disease have few therapeutic alternatives. Because the existing data suggest that the toxicity of the purine antimetabolites is not likely to be a major problem, AZA can be used in patients with poor prognosis despite some lingering concerns about efficacy.

Brian Feagan, MD
University of Western OntarioLondon, Ontario, Canada


1. Singleton JW. Azathioprine has a very limited role in the treatment of Crohn's disease. Dig Dis Sci. 1981;26:368-71.

2. Korelitz BI, Present DH. Shortcomings of the National Crohn's Disease Study: the exclusion of azathioprine without adequate trial. Gastroenterology. 1981;80:193-6.