Search

When dealing with missing data in clinical trials, it is often convenient to work under simplifying assumptions, such as missing at random (MAR), and follow up with sensitivity analyses to address unverifiable missing data assumptions. One such sensitivity analysis, routinely requested by regulatory agencies, is the so-called tipping point analysis, in which the treatment effect is re-evaluated after adding a successively more extreme shift parameter to the predicted values among subjects with missing data. If the shift parameter needed to overturn the conclusion is so extreme that it is considered clinically implausible, then this indicates robustness to missing data assumptions. Tipping point analyses are frequently used in the context of continuous outcome data under multiple imputation. While simple to implement, computation can be cumbersome in the two-way setting where both comparator and active arms are shifted, essentially requiring the evaluation of a two-dimensional grid of models. We describe a computationally efficient approach to performing two-way tipping point analysis in the setting of continuous outcome data with multiple imputation. We show how geometric properties can lead to further simplification when exploring the impact of missing data. Lastly, we propose a novel extension to a multi-way setting which yields simple and general sufficient conditions for robustness to missing data assumptions.

Objective: To investigate the efficacy, safety, and tolerability of oral semaglutide added to insulin with or without metformin., Research Design and Methods: Patients with type 2 diabetes uncontrolled on insulin with or without metformin were randomized to oral semaglutide 3 mg ( N = 184), 7 mg ( N = 182), or 14 mg ( N = 181) or to placebo ( N = 184) in a 52-week, double-blind trial. End points were change from baseline to week 26 in HbA 1c (primary) and body weight (confirmatory secondary). Two estimands were defined: treatment policy (effect regardless of trial product discontinuation or rescue medication) and trial product (effect assuming trial product continuation without rescue medication) in randomized patients., Results: Oral semaglutide was superior to placebo in reducing HbA 1c (estimated treatment difference [ETD] -0.5% [95% CI -0.7, -0.3], -0.9% [-1.1, -0.7], and -1.2% [-1.4, -1.0] for 3, 7, and 14 mg, respectively; P < 0.001) and body weight (ETD -0.9 kg [95% CI -1.8, -0.0], -2.0 kg [-3.0, -1.0], and -3.3 kg [-4.2, -2.3]; P = 0.0392 for 3 mg, P ≤ 0.0001 for 7 and 14 mg) at week 26 (treatment policy estimand). Significantly greater dose-dependent HbA 1c and body weight reductions versus placebo were achieved with oral semaglutide at weeks 26 and 52 (both estimands). The most frequent adverse event with oral semaglutide was nausea (11.4-23.2% of patients vs. 7.1% with placebo; mostly mild to moderate)., Conclusions: Oral semaglutide was superior to placebo in reducing HbA 1c and body weight when added to insulin with or without metformin in patients with type 2 diabetes. The safety profile was consistent with other glucagon-like peptide 1 receptor agonists., (© 2019 by the American Diabetes Association.)

Importance: Achieving glycemic control remains a challenge for patients with type 2 diabetes, even with insulin therapy., Objective: To assess whether a fixed ratio of insulin degludec/liraglutide was noninferior to continued titration of insulin glargine in patients with uncontrolled type 2 diabetes treated with insulin glargine and metformin., Design, Setting, and Participants: Phase 3, multinational, multicenter, 26-week, randomized, open-label, 2-group, treat-to-target trial conducted at 75 centers in 10 countries from September 2013 to November 2014 among 557 patients with uncontrolled diabetes treated with glargine (20-50 U) and metformin (≥1500 mg/d) with glycated hemoglobin (HbA1c) levels of 7% to 10% and a body mass index of 40 or lower., Interventions: 1:1 randomization to degludec/liraglutide (n = 278; maximum dose, 50 U of degludec/1.8 mg of liraglutide) or glargine (n = 279; no maximum dose), with twice-weekly titration to a glucose target of 72 to 90 mg/dL., Main Outcomes and Measures: Primary outcome measure was change in HbA1c level after 26 weeks, with a noninferiority margin of 0.3% (upper bound of 95% CI, <0.3%). If noninferiority of degludec/liraglutide was achieved, secondary end points were tested for statistical superiority and included change in HbA1c level, change in body weight, and rate of confirmed hypoglycemic episodes., Results: Among 557 randomized patients (mean: age, 58.8 years; women, 49.7%), 92.5% of patients completed the trial and provided data at 26 weeks. Baseline HbA1c level was 8.4% for the degludec/liraglutide group and 8.2% for the glargine group. HbA1c level reduction was greater with degludec/liraglutide vs glargine (-1.81% for the degludec/liraglutide group vs -1.13% for the glargine group; estimated treatment difference [ETD], -0.59% [95% CI, -0.74% to -0.45%]), meeting criteria for noninferiority (P < .001), and also meeting criteria for statistical superiority (P < .001). Treatment with degludec/liraglutide was also associated with weight loss compared with weight gain with glargine (-1.4 kg for degludec/liraglutide vs 1.8 kg for glargine; ETD, -3.20 kg [95% CI, -3.77 to -2.64],P < .001) and fewer confirmed hypoglycemic episodes (episodes/patient-year exposure, 2.23 for degludec/liraglutide vs 5.05 for glargine; estimated rate ratio, 0.43 [95% CI, 0.30 to 0.61],P < .001). Overall and serious adverse event rates were similar in the 2 groups, except for more nonserious gastrointestinal adverse events reported with degludec/liraglutide (adverse events, 79 for degludec/liraglutide vs 18 for glargine)., Conclusions and Relevance: Among patients with uncontrolled type 2 diabetes taking glargine and metformin, treatment with degludec/liraglutide compared with up-titration of glargine resulted in noninferior HbA1c levels, with secondary analyses indicating greater HbA1c level reduction after 26 weeks of treatment. Further studies are needed to assess longer-term efficacy and safety., Trial Registration: clinicaltrials.gov Identifier: NCT01952145.