Use of Prohibited Medication, a Potentially Overlooked Confounder in Clinical Trials: Omarigliptin (Once-weekly DPP-4 Inhibitor) Monotherapy Trial in 18- to 45-year-olds.

PURPOSE: The objective of this clinical trial was to assess the efficacy and safety of omarigliptin monotherapy in young adult patients with type 2 diabetes mellitus (T2DM). Unexpected efficacy results in this trial led to a series of investigations that identified the use of prohibited medication by a substantial number of trial patients. METHODS: Patients with T2DM who were ≥18 to <45 years of age and either drug-naive or not on an antihyperglycemic agent for ≥12 weeks with inadequate glycemic control were randomized in a double-blind manner to receive omarigliptin 25 mg once weekly (n = 102) or placebo once weekly (n = 101) for 24 weeks. The objectives of the trial were to assess the effect of treatment with omarigliptin on glycemic parameters, including levels of glycosylated hemoglobin (HbA1c), 2-hour postmeal glucose, and fasting plasma glucose, and to assess the safety and tolerability of omarigliptin. Additional investigations into trial conduct included the measurement of drug levels for omarigliptin and metformin in blood samples collected for future biomedical research, available for approximately one half of the patients. FINDINGS: The mean age of trial participants was 39.2 years, approximately 60% were male, mean body mass index was 32.5 kg/m2, and mean duration of diabetes was 3.1 years. The mean baseline HbA1c value was 7.9% in the omarigliptin group and 8.1% in the placebo group. After 24 weeks, the least squares mean change (95% CI) in HbA1c value from baseline was -0.33% (-0.60 to -0.06) in the omarigliptin group and -0.45% (-0.72 to -0.18) in the placebo group, with a between-group difference of 0.12% (-0.26 to 0.49; P = 0.535). Similarly, no between-group difference was observed for the other glycemic parameters (2-hour postmeal glucose and fasting plasma glucose levels). No issues were identified in drug allocation, dispensing or supply, patient compliance with trial medication, sample handling or analysis, or site trial conduct that explained the observed results. Measurement of drug levels from future biomedical research samples uncovered the use, with no investigator knowledge, of an antihyperglycemic agent that was prohibited by the protocol (ie, metformin) by 42.4% (39 of 92) of patients. Metformin was used by more patients in the placebo group (57% [25 of 44]) than in the omarigliptin group (29% [14 of 48]). IMPLICATIONS: The use of prohibited metformin in a trial of a dipeptidyl peptidase-4 inhibitor, omarigliptin, introduced a confounding factor that invalidated the results of the trial. This behavior may have been encouraged in the trial by protocol-specified self-monitoring of blood glucose levels. Use of prohibited medication may be an underappreciated confounder in clinical trial research. TRIAL REGISTRATIONS: MK-3102-028 (US); ClinicalTrials.gov identifier, NCT01814748; EudraCT number, 2012-004303-12 (EU).