Head‐to‐head trials in inflammatory bowel disease

Abstract

Head-to-head trials are typically either classified as non-inferiority, equivalence, or superiority studies. The decision as to which of these to choose is complex and requires an in-depth knowledge of existing studies and data that pertain to the existing agent and the (new) comparator. Recent trials of biologics for IBD have used either a non-inferiority or a superiority study design. For the former, typically a newer agent is compared with an existing, established agent. The aim is to determine whether the new agent is no less effective than the comparator. These agents may have similar efficacies but may differ in route of administration, cost, and safety profile. A major methodological step is in the choice of the non-inferiority margin, which is the largest acceptable loss of efficacy of the new drug relative to the comparator. This margin will influence the sample size required for the study and hence other major issues such as recruitment time and costs. The narrower this margin, the greater the sample size required to achieve the primary objective. A recent example of a non-inferiority trial is the NOR-SWITCH study comparing the efficacy and safety of the originator infliximab (Remicade®) with a biosimilar infliximab (CT-P13). Given the small population within Norway, this study selected participants across multiple indications and analyzed the data as an aggregate response across all indications using validated measures of response. The non-inferiority margin was set at 15% which is in line with European Medicines Agency (EMA) requirements. As such power calculations dictated that 197 participants were required in each group. While the primary outcome was achieved, that is, non-inferiority was demonstrated across the aggregate of six indications (risk difference, 4.4%), the Crohn’s disease subgroup was very close to demonstrating inferiority for the biosimilar (risk difference, 14.3%). Positive features of this study included the clinical outcomes, government funding, and the collection of additional data on drug levels and anti-drug antibody status. Unfortunately, it was not powered to examine individual indications. Overall, it was clinically meaningful and at the time, up to date. Superiority trials investigating biologics have been rare until recently. Initially commencing in rheumatology and dermatology, superiority trials now are being undertaken in patients with IBD. This is particularly important given the increasing number of agents available to both patients and clinicians. These studies have been eagerly awaited given the ethical concerns about placebo arms in studies of patients with moderate to severe IBD activity who have either failed or had an inadequate response to existing (historical) therapies. It is critical that study design is relevant to both current and near to medium-term future clinical practice. Such design features include robust and validated measures of disease activity, dosing, therapeutic-drug monitoring, concomitant IBD therapy, and effective strategies to achieve timely subject recruitment. A recent example of a randomized, double-blind, superiority trial is the VARSITY study comparing vedolizumab with adalimumab for moderate to severe ulcerative colitis (UC). The trial commenced recruitment in July 2015, involved 245 sites globally, and was sponsored by Takeda. The primary endpoint was clinical remission at 52 weeks while secondary endpoints included corticosteroid-free clinical remission at 52 weeks and endoscopic improvement. An important aspect of study design perhaps related to potential recruitment challenges was the decision to include bio-experienced patients (anti-tumor necrosis factor [TNF] other than adalimumab) in the study (N = 769). Vedolizumab (31.3%) demonstrated superior clinical remission to adalimumab (22.5%, P = 0.006) at 12 months, with a stronger signal for vedolizumab in the anti-TNF naïve subgroup as compared to the anti-TNF experienced subgroup. Vedolizumab also achieved superiority for endoscopic improvement (39.7% versus 27.7%, P < 0.001) while adalimumab achieved a higher rate of corticosteroid-free remission (21.8%, versus vedolizumab, 12.6%, 95% confidence interval [CI], 18.9 to 0.4). Dose adjustment and drug levels/anti-drug antibodies were not included in the study design. Sample size estimates were based upon remission rates at 52 weeks of 28% for vedolizumab and 18% for adalimumab. Other head-to-head trials that are underway and/or nearing completion include trials of etrolizumab versus infliximab and vedolizumab versus infliximab in moderate to severe UC. Although results are keenly awaited, there may well be significant challenges in recruitment and control of variance in care/data quality across hundreds of sites globally. Alternatives to these very costly prospective trials do exist including studies that are embedded in current, established clinical practice, and using techniques such as comparative effectiveness research and propensity score matching. An example of this is the REACT study which compared early combined immunosuppression versus conventional management in Crohn’s disease. In summary, head-to-head biologic studies in IBD are a major step forward in determining optimal therapeutic strategies for patients with moderate to severe disease. Study design and the use of additional or alternative strategies to mitigate against challenging recruitment targets, costs, and changes in treatment paradigms are important factors for current and future investigators, sponsors, and government bodies.