Smoking and Early Infliximab Response in Crohn's Disease: a Meta-analysis

Background: Infliximab is used to treat moderate to severe Crohn's disease (CD), but its efficacy varies. Although cigarette smoking worsens CD, its impact on the infliximab response is unknown. We conducted a systematic review and meta-analysis of clinical trials to determine the effect of smoking on the induction response to infliximab. Methods: A systematic search was performed of MEDLINE, EMBASE, CINAHL, the Cochrane central register of controlled trials, the Cochrane IBD Group Specialized Trials Register for publications, and abstracts from major conferences from January 1996 to December 2010. Random effects meta-analysis using the Mantel-Haenszel method was conducted. Heterogeneity across studies was assessed using the Q statistic, the I-2 statistic, and tau(2). Results: We identified 12 articles; four were excluded due to use of non-validated scoring systems. The remaining eight included a total of 1658 patients, with 649 active smokers. Luminal response was assessed by the Crohn's Disease Activity Index in four studies (three of which included fistula response) and the Harvey-Bradshaw index in two (both including fistula response), and two studies examined only the fistula response. The relative risk for response to infliximab among smokers was 0.99 (95% CI 0.88-1.11) (tau(2) = 0.0143). Analyses of the five studies examining both inflammatory and fistulizing CD were similar to the analysis of all eight studies. The pooled relative risk was 0.92 (95% CI 0.80-1.06) (tau(2) = 0.0154). Conclusion: Though smoking worsens CD, this meta-analysis does not show a negative effect of smoking on initial response to infliximab. This must be viewed in the proper context, as long-term maintenance of response may yet be influenced by smoking status.