Association between 6-thioguanine nucleotides levels and clinical remission in inflammatory disease: a meta-analysis

A previous meta-analysis suggested that 6-thioguanine nucleotides levels are associated with clinical remission in inflammatory bowel disease. It was criticized because of the relatively small number of patients included in this meta-analysis and heterogeneity between studies. Recent studies provided conflicting results, and the source of those discrepancies has yet to be explored. A comprehensive, computerized literature search was conducted in Medline, ISI Web of Science, and EMBASE until December 31, 2012. A combined odd ratio with its 95% confidence interval was calculated using a fixed effects model based on the Mantel-Haenszel method. Between-study heterogeneity was assessed using Cochran's Q statistic. Seventeen studies enrolling 2049 patients with inflammatory bowel disease were analyzed. A significant heterogeneity was found in the overall analysis (P = 0.005). As heterogeneity among studies could be explained by differences in metabolite assay methods, an analysis including only studies using the reference method by Lennard et al (N = 10) was performed, and the pooled odds ratio for clinical remission among patients with 6-thioguanine nucleotides levels over a cut-off value between 230 and 260 pmol/8.10^8 RBC was 3.15 (95% confidence interval, 2.41-4.11). This meta-analysis clearly establishes an association between 6-thioguanine nucleotides levels and clinical remission rates in patients with inflammatory bowel disease and explains the heterogeneity of results among selected studies. The lack of standardization in 6-thioguanine nucleotides assays is responsible for recent contradictory results. Whether therapeutic drug monitoring of thiopurines should be systematically used in clinical practice in inflammatory bowel disease to improve disease outcomes will require further investigation.