Research in context
Evidence before this study
We searched PubMed from inception to May 31, 2018, for previously published meta-analyses on pharmacological interventions for mild to moderate ulcerative colitis using the search terms “ulcerative colitis” and “treatment” without any language restrictions. Our search identified 869 systematic reviews, but no network meta-analyses. Previous traditional pairwise meta-analyses provide a limited synthesis of comparative efficacy of different interventions, limited to head-to-head trials, without incorporating indirect evidence. Additionally, the quality of this evidence has not been critically examined, which might inform clinical guidelines.
Added value of this study
Our study provides the first network meta-analysis of all pharmacological interventions for management of mild to moderate ulcerative colitis, with systematic assessment of quality of the evidence using Grading of Recommendations Assessment, Development and Evaluation methodology. Through a systematic review with network meta-analysis, including 48 randomised trials (8020 participants) comparing eight active interventions for induction of remission, and 28 randomised trials (4218 participants) comparing six active interventions for maintenance of remission in patients with left-sided or extensive ulcerative colitis, we are able to inform the treatment approach in these patients. For induction of remission, a combined oral and rectal 5-aminosalicylate (ASA) is the most effective treatment strategy, followed by high-dose mesalazine (moderate confidence in estimates); standard-dose mesalazine is similar to diazo-bonded 5-ASAs, but superior to sulfasalazine. For maintenance of remission, standard-dose and high-dose mesalazine are similar (low confidence in estimates), and superior to low-dose mesalazine (high confidence in estimates). Budesonide multimatrix is not more effective than a combined oral and rectal 5-ASA or high-dose mesalazine and has inferior tolerability.
Implications of all the available evidence
Contextualising findings on comparative efficacy and tolerability, coupled with patients' values and preference, and cost and resource use considerations, our findings can directly inform treatment guidelines. Patients with mild ulcerative colitis are likely to achieve remission with standard-dose mesalazine or a diazo-bonded 5-ASA. A subset of patients, particularly those with more moderately active disease, or who live in areas where high-unit dose forms of mesalazine are available at a low cost, physicians might reasonably decide to use high-dose oral mesalazine for induction of remission. In patients with extensive colitis, a rectal 5-ASA might be added either upfront to standard-dose or high-dose oral mesalazine, or in 4–6 weeks in a subset of patients with suboptimal response to oral therapy. Likewise, budesonide multimatrix might be considered as monotherapy or in addition to 5-ASA therapy in patients with moderately active disease, either upfront or in the case of suboptimal response to optimised 5-ASA therapy. For maintenance of remission, standard-dose mesalazine and a diazo-bonded 5-ASA might be preferred options. High-dose mesalazine with or without a rectal 5-ASA is no more effective than standard-dose mesalazine, and consideration should be made to cautiously de-escalate patients who required high-dose mesalazine for induction of remission to standard dose, after 6–12 months.