Investigation of bioequivalence of a new fixed‐dose combination of acarbose and metformin with the corresponding loose combination as well as the drug–drug interaction potential between both drugs in healthy adult male subjects

Summary What is known and objective Both metformin and acarbose are recommended monotherapy and add‐on therapy in type 2 diabetes mellitus ( T 2 DM ). A fixed‐dose combination ( FDC ) of acarbose and metformin has been developed to reduce pill burden and potentially improve compliance. The current study investigated the bioequivalence of the acarbose/metformin FDC compared with the individual agents administered simultaneously (loose combination). Secondary endpoints were the safety and tolerability of the FDC and the potential for drug–drug interactions between acarbose and metformin. Methods A single‐centre, randomized, open‐label, four‐period crossover study was conducted in healthy male K orean subjects aged 18–45 years. Following one‐period balanced Williams design, participants were randomized to receive four single oral treatments on different study days separated by ≥7 days' washout. Treatments were as follows: (i) acarbose/metformin 50/500 mg FDC (test); (ii) acarbose 50 mg and metformin 500 mg as loose combination (reference); (iii) acarbose 50 mg; and (iv) metformin 500 mg. Serial blood samples were taken for glucose and insulin levels for 4 h after a sucrose load on the day before and day of study drug administration. Additionally, serial blood samples were taken for analysis of metformin levels for 24 h after each drug containing metformin. The area under the curve for 4 h post‐test ( AUC 0–4 h ) and the maximal serum concentration ( C max ) of plasma glucose and serum insulin were primary pharmacodynamic ( PD ) parameters, and C max , AUC 0–last and AUC for metformin levels were primary pharmacokinetic ( PK ) parameters. The bioequivalence of the FDC to the loose combination was considered established if the 90% confidence intervals ( CI s) of the baseline‐adjusted PD parameter ratios (test vs. reference) for plasma glucose and the PK parameter ratios for metformin fell completely within current acceptance limits (0·8–1·25). Results and discussion Thirty‐three of 40 randomized subjects completed the study; five withdrew consent and two discontinued because of adverse events ( AE s). The 24‐h plasma concentration–time curves of metformin and the 4‐h plasma glucose–time curves after acarbose/metformin FDC (test) and acarbose + metformin loose combination (reference) were almost superimposable. The geometric least squares ( LS ) mean of the Ratio AUC and Ratio C maxfor plasma glucose after the FDC vs. loose combination, and the LS mean of the ratios in metformin AUC , AUC 0–last and C max were close to unity, and the 90% CI of all these parameters fell within the predefined equivalence range of 0·8–1·25, confirming bioequivalence. The metformin AUC was reduced by 26% and C max by 34% after acarbose + metformin compared with metformin alone. Eight subjects (20·0%) reported AE s, but all were mild, and most were gastrointestinal, as expected for these agents. The incidence of AE s was not higher with the combinations vs. monotherapy. What is new and conclusion These data demonstrate that the acarbose/metformin FDC is bioequivalent to the loose combination of these agents. Although acarbose slightly reduced the bioavailability of metformin, the accumulated evidence of the efficacy of this combination implies that this is clinically irrelevant. The observed AE profile was consistent with the established knowledge on the safety of the two drugs.