Is Mesalamine Effective for the Induction of Remission in Crohn's Disease?
We read with interest the Bayesian network meta-analysis by Coward et al,1 comparing the effectiveness of mesalamine, sulfasalazine, corticosteroids, and budesonide for the induction of remission in Crohn's disease. They found high-dose mesalamine (i.e., at doses ≥2.4 g/d) to be more effective than placebo (odds ratio = 2.29; 95% credible interval: 1.58–3.33) and conclude that high-dose mesalamine is an option among patients preferring to avoid steroids.1
This statement conflicts with current clinical practice guidelines from the American College of Gastroenterology, which consider mesalamine as minimally effective for the treatment of mild-to-moderate Crohn's disease.2 This unexpected finding is also inconsistent with the results of our recent network meta-analysis3 and a previous Cochrane systematic review,4 which both show a lack of effectiveness of mesalamine as induction agent in Crohn's disease, although all 3 meta-analyses have included the same mesalamine studies.
After careful assessment of the published manuscript, we must caution that a data extraction error with regards to the Crohn's II study has seriously affected the mesalamine-specific effect estimates in the meta-analysis by Coward et al.1 The data for this large, randomized controlled trial were incorrectly extracted from a previous review (Ford et al,5 Figure 2). In their analysis, Coward et al1 included 117/157 (75%) remissions in the mesalamine group and 51/75 (68%) remissions among the placebo controls (Coward et al,1 Table 1, Crohn's II), whereas the correct numbers are 40/157 (25%) and 24/75 (32%); thus, they erroneously turned Crohn's II study from negative (i.e., in favor of placebo) to positive (in favor of mesalamine). Unfortunately, the authors did not notice the “reverse” definition of events in this review; Ford et al5 defined as “event” the failure to achieve remission in trials of therapy for active Crohn's disease. Accordingly, relative risks higher than 1.0 favor placebo, whereas relative risks lower than 1.0 favor mesalamine in the respective forest plot (Ford et al,5 Figure 2).
It is the reason why we strongly believe that their findings should be reassessed after correct data extraction. The authors should then be able to confirm the lack of effectiveness of mesalamine for the induction of remission in Crohn's disease because the analysis incorporates the same mesalamine studies with previous works.