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


Review: Most current drugs improve active Crohn disease but do not prevent relapse

ACP J Club. 1992 Nov-Dec;117:69. doi:10.7326/ACPJC-1992-117-3-069

Source Citation

Salomon P, Kornbluth A, Aisenberg J, Janowitz HD. How effective are current drugs for Crohn's disease? A meta-analysis. J Clin Gastroenterol. 1992 Mar;14:211-5.



To evaluate the effectiveness of drugs currently in use for the treatment of Crohn disease in inducing and maintaining remission.

Data sources

A MEDLINE search was done for the period 1967 through May 1991 using the terms Crohn disease, controlled, trials, and placebo. Additional published studies were identified from bibliographies.

Study selection

12 full-text studies of single drugs compared with placebo that had well-defined study duration and end points were selected by blinded review of methods sections. Trials of drugs not currently in use and those allowing concurrent therapy were excluded.

Data extraction

Drugs used (sulfasalazine [SAS], azathioprine [AZA], steroids, 5-aminosalicylic acid [5-ASA], and cotrimazole and metronidazole); details of study quality, size, and duration; patients' ages and duration of disease; and percentages of patients achieving and maintaining complete clinical remission or partial improvement were extracted. Criteria for remission were a score of < 150 on the Crohn's Disease Activity Index (6 trials), < 140 on the van Hees index (1 trial), and comparable scores for other indexes. Study dropouts were considered treatment failures in intention-to-treat analysis. The authors pooled the results of 7 trials for the induction of remission of active disease and 5 trials for the maintenance of remission.

Main results

A therapeutic advantage (the rate of response to a drug minus the rate on placebo) of the drugs of 10% to 29% was achieved for complete clinical remission (absolute risk difference [RD] between drugs and placebo 20%, 95% CI 9% to 31%). For partial clinical improvement, therapeutic advantage was 10% to 34% (RD 23%, CI 8% to 37%). SAS and AZA had low therapeutic advantage over placebo in inducing complete remission (SAS: RD 13%, CI 3% to 24%, and AZA: RD 10%, CI 6% to 25%). Steroids had a greater therapeutic advantage (RD 33%, CI 9% to 56%). In the trials reporting patients with complete or partial improvement (variously defined), the therapeutic advantage of SAS was greatest (RD 38%) compared with 5-ASA (RD 10%) and antibiotics (RD 17%).

No drug maintained remission better than placebo; relapse rate to 24 or 36 months was approximately 90% for patients assigned to active treatment or placebo.


Drugs currently used for the treatment of Crohn disease effectively promote clinical improvement and complete remission but do not prevent relapse during the following 3 years.

Source of funding: Not stated.

For article reprint: Dr. P. Salomon, The Mount Sinai Hospital, 1 Gustave Levy Place, Annenberg Building, 23rd Floor, New York, NY 10029, USA.


Olsalazine was more effective than mesalazine in maintaining remission from ulcerative colitis

The 2 studies address many of the questions raised by clinical trials in inflammatory bowel disease. The article by Courtney and colleagues describes a large single-center comparative study of 2 newer aminosalicylates for the maintenance of remission in ulcerative colitis. The methods are similar to those used in previous trials comparing ASA derivatives with sulfasalazine, which found comparable efficacy but improved tolerability of the ASA derivatives compared with the standard (sulfasalazine) treatment. The results show that the azo-linked olsalazine compound, which delivers 5-ASA to the colon in a manner similar to sulfasalazine, is superior to the acrylic-coated preparation, mesalazine (Asacol).

Interpretation of the results should, however, take several factors into account. To optimize compliance (which was checked) the patients knew which drug they were taking; therefore previous experience with one or both of the medications at entry to the trial may have influenced their responses to treatment. Moreover, although the investigators were blinded to treatment, their definition of "remission" and "relapse" makes it difficult to completely exclude a systematic bias. It would also be helpful for clinicians to know whether different doses of the 2 drugs would have achieved similar results. In this study the drugs were administered at the manufacturers' recommended dosage; olsalazine was administered as 4 tablets, 250 mg each, and mesalazine as 3 tablets, 400 mg each. Although the compliance of the 2 groups of patients was comparable, the differences in results could be explained by a lack of comparability between the mesalazine and olsalazine delivery systems. The effectiveness might have been similar if higher doses of mesalazine (e.g., 4 tablets, 1.6 g) had been used.

Salomon and colleagues used the technique of meta-analysis to determine the overall efficacy of drugs currently used for the treatment of Crohn disease. The authors emphasize that different end points and patient populations were included that may obscure effectiveness or lack thereof in subgroups of patients with Crohn disease. Nevertheless, there is a consistency in clinical trials showing a relatively modest benefit for our current therapies, which include sulfasalazine, 5-ASA, steroids, and immunosuppressives. The "therapeutic advantage" in inducing complete clinical remission, which ranged from 10% for azathioprine and 11% for sulfasalazine to 29% for steroids in active Crohn disease, emphasizes the need to continue exploring new therapeutic approaches. The meta-analysis includes only trials that used single agents, whereas clinicians recognize the frequent empiric need for combination approaches. Future studies will need both to assess single and combination therapies in predefined subsets of patients with Crohn disease and to have well-defined clinical end points.

The 2 studies emphasize that we can no longer lump all patients with ulcerative colitis or Crohn disease together, treat them with a single drug at a single dose, and expect to identify substantial differences that make clinical practice straightforward. Until larger trials incorporating predefined subpopulations of patients treated with varying doses or combinations are done, we will have to depend on results from limited controlled trials combined with clinical experience to form individualized approaches to patients. Although this has been the basis for our therapeutic decisions, we hope and anticipate that clinical research in inflammatory bowel disease will continue to advance as our ability to distinguish clinically important differences in patients improves.

Stephen B. Hanauer, MD
University of Chicago Medical CenterChicago, Illinois, USA