How to Accurately Establish Pharmacokinetics/Pharmacodynamics of Long-Acting Insulins in Humans: Relevance to Biosimilar Insulins.

In this issue of Diabetes Care, Linnebjerg et al. (1) compare the pharmacokinetics (PK) and pharmacodynamics (PD) of a Lilly synthetized insulin glargine (LY IGlar) with the insulin glargine Lantus (IGlar) based on a series of clamp studies in healthy volunteers and conclude for PK/PD that there is similarity of the LY IGlar vs. IGlar. The strengths of the study should be discussed along with its limitations. The strengths of the study are several. First, there are consistent rate ratios of geometric means of the several PK/PD parameters of LY IGlar and IGlar examined, which are all close to 1.0, with the 90% CIs contained in the prespecified acceptance limits of 0.80–1.25. Second, there is an elevated number of subjects studied with the technique of euglycemic clamp (N = 211) and the study design (randomized, double-blind, two-treatment, four periods, crossover) is strong. Third, the comparison between LY IGlar and IGlar is repeated twice (one time vs. European Union [EU]-approved IGlar, the other time vs. US-approved IGlar), and there is the unique comparison between EU-approved IGlar and US-approved IGlar. Of note, the authors not only demonstrate PK/PD similarity between LY IGlar and IGlar but also originally prove the intraindividual similarity (reproducibility) of different batches of IGlar (EUand US-approved IGlar both originate from the same manufacturer, Sanofi in Frankfurt, Germany). The limitations of the study by Linnebjerg et al. (1) derive from study design and population studied. The conclusions are correct in the specific conditions examined in this study, i.e., normal volunteers without diabetes, a single insulin injection, a dose of 0.5 units/kg (supra-therapeutic, at least for subjects with type 1 diabetes [T1D]), and an insulin dose given in the morning. It is our opinion that one should be cautious regarding the resultswhenextrapolating to the general population of “users” of basal insulin, i.e., subjects with T1D or type 2 diabetes (T2D) who inject basal insulin every day (at steady state), usually at lower doses and nearly always in the evening. The study by Linnebjerg et al. (1) opens a number of interesting and relevant questions. How can we best establish PK/PD of long-acting insulins, which today have duration of action beyond 24 h?Which is the correct methodology of the glucose clamp technique and how should it be used? In which category subjects should the clamp studies be done? The premise is that the clamp studies should be designed and executed not only to fulfill the requests of the regulatory agencies (2,3) but also to provide meaningful and objective evidence for doctors and patients on how the new candidate basal insulins work in clinical practice as compared with standard treatment. To be as meaningful as possible, the comparison should be done in experimental conditions as close to the real life of persons with diabetes.