• Crohn's disease
• azathioprine
• 5-aminosalicylic acid
• surgery
• relapse
• surgery for IBD

I read the study performed by Reinisch and colleagues with interest.1 As the authors state, the optimum strategy for the treatment of Crohn's disease following intestinal resection remains uncertain, with conflicting evidence for use of mesalazine,2 3 and only limited data to suggest that 6-mercaptopurine (6-MP) or azathioprine are more effective than placebo4 5 in terms of the prevention of either clinical or endoscopic relapse of disease activity.

Reinisch et al are therefore to be commended for conducting the first randomised double-blind placebo-controlled trial to compare mesalazine and azathioprine in this situation. The results of the study failed to demonstrate inferiority of mesalazine compared with azathioprine, in terms of the rates of therapeutic failure (defined as clinical recurrence or discontinuation of treatment due to lack of efficacy or side effects), which were 11% and 22%, respectively.

What is unclear is why the power calculation that the authors used assumed a difference in treatment effect of 35% in favour of azathioprine. There have been two previous randomised controlled trials that have compared the use of 6-MP or azathioprine with mesalazine.4 6 These demonstrated a difference in efficacy somewhat short of this magnitude, 8% in favour of 6-MP and 11% in favour of azathioprine. It should be noted, however, that these two trials included a subtly different trial population to the one in the present study, in that endoscopic recurrence of disease activity following resection was not confirmed prior to randomisation. If the power calculation were to be redone using the difference in efficacy observed in the previous azathioprine study (11%) and the pooled therapeutic failure rate observed in placebo-controlled trials of mesalazine in postoperative Crohn's disease (37%, author's own data) then the sample size required in each arm at the 80% power and 5% significance level would be 280 patients.

As the investigators correctly point out in their discussion, one of the limitations of the study may be that the efficacy of mesalazine was underestimated. However, it is also probable that the superiority of azathioprine over mesalazine in this situation was overestimated, leading to a lack of power of the study to detect a statistically significant difference in therapeutic efficacy between the two drugs.