Patent protection of several recombinant insulins has expired or will expire within the near future, providing the opportunity to the pharmaceutical industry to develop biosimilar insulins for competition with the marketing authorisation holder of the originator insulin. Biosimilar Insulin is generally defined as a biological medicine developed to be similar to a biological medicine already approved for human use. While generics are considered to be chemically identical to their reference product, biosimilars are regarded to be different from the reference product. Biosimilar peptides distinguish from chemical generics with regard to the complexity of the production process and the regulatory requirements for the approval of the drug in several countries (1). Biosimilar peptides feature the same primary amino acid sequence as the reference product, but minor differences in the production process might cause clinically significant deviations in the biological behaviour of the peptide. The synthesis of recombinant insulin is a highly complex procedure including several different processing steps. Insulin synthesis relies on the insertion of the desired nucleotide sequence, coding for the precursor, pro-insulin, into appropriate host cells. After fermentation, the achieved pro-insulin has to be cleaved to insulin and c-peptide, followed by a purification of the insulin using high resolution chromatography. Any variation in the production process like variation in cell lines, media components, expression systems, bioreactor conditions, or the purification process might significantly affect the clinical characteristics of the final product especially with regard to its pharmacokinetic and pharmacodynamic properties, or the immunogenic potency. Small variations in tertiary peptide folding or the arrangement of insulin molecules to one another might further modify insulin kinetics or insulin receptor interactions.
Thus, before biosimilar insulin receives market approval, it has to ensure that it is safe and has a comparable pharmacokinetic and pharmacodynamics profile as the reference insulin. Several guidelines from regulatory authorities have been published to guide through the regulatory process of market approval for biosimilar insulins in different regions of the world. Areas covered by these guidelines include preclinical and clinical studies assuring safety and efficacy of the biosimilar insulin. Even if the guidelines are inherently comparable, emphasis of different aspects might diverge to some extend in distinct countries. To receive approval of a new biosimilar, the applicant has to provide documentation of the full manufacturing process, the active ingredient including molecular structures, degradation products, as well as each component included in the final formulation. Even though a biosimilar needs to have an identical primary amino acid sequence, appropriate physico-chemical tests should demonstrate comparable primary, secondary, and tertiary structures of the molecule. Beyond demonstrating identical physico-chemical properties, predefined toxicological and clinical studies are required. In vitro binding assays are required to demonstrate comparable insulin (IR) and IGF-1 receptor affinities and intracellular signalling. Metabolic and mitogenic potency should be addressed providing sufficient animal and cell culture experiments. It is not requested that the biosimilar insulin is identical to the reference insulin in terms of formulation or excipients, but every change in the final formulation has to be justified.
Treatment with insulin requires a very sensitive dose titration, and already small deviations in the pharmacokinetic or pharmacodynamic profile of insulin might implicate clinically critical glucose excursion. The time-concentration (pharmacokinetic) and time action (pharmacodynamic) profiles are considered to be the mainstay for the proof of similarity between biosimilar insulin and the reference insulin. The hyperinsulinaemic, euglycaemic/isoglycaemic clamp technology is considered to be the gold standard for the clinical characterisation of investigational insulins. In this experiment, a constant rate of insulin is infused to supress endogenous glucose production. After subcutaneous application of the test insulin, the time profile of the glucose infusion rate (GIR) to keep the blood glucose in the euglycaemic/isoglycaemic range is recorded for the characterisation of the pharmacodynamic properties of insulin. Additional measurement of plasma insulin concentrations are used to demonstrate the pharmacodynamic profile. Different clamp technologies and feedback algorithms have been developed for the adaptation of the glucose infusion rate. Clamp studies can be performed manually or by the use of an automated computerized clamp system. Automated clamp systems provide the advantage of more frequent blood glucose measurements (every minute), tight adaptation of glucose infusion rates, and less investigator bias. New automated glucose clamp systems have been implemented to reduce the variability and to improve the reproducibility of the clamp experiments (2). In any way, these clamps should be performed at experienced study sites.
The European Medical Agency (EMA) provides particular detailed guidance on clamp methodology, endpoints, and statistical analysis for the registration of biosimilar insulin’s (3).
According to the EMA guidelines, the study population should consist of normal weight healthy volunteers or patients with type 1 diabetes mellitus. In type 1 diabetic patients, serum c-peptide levels should be measured to exclude residual insulin secretion from the beta cell. It is recommended to include only male subjects to exclude potential effects of the menstrual cycle on insulin sensitivity. The dose of subcutaneously applied insulin should mimic commonly used therapeutic ranges, and the clamp duration should be planed according to the pharmacokinetic properties of the investigational insulin. Clamp durations of eight to ten hours for rapid insulin and up to forty-eight hours for long acting insulin might be required to reflect the pharmacokinetic profile of the insulin.
Comprehensive comparative data on the time –concentration profile including AUC and Cmax as the primary, and Tmax, early and late T50%, and T1/2 as secondary pharmacokinetic endpoints should be calculated. For AUC and Cmax the 90% confidence interval of the ratio test/reference insulin should lie within 80 – 125 %. The glucose infusion rate (GIR) over time characterizes the pharmacodynamic properties of the insulin molecule. GIRAUC and GIRmax serve as primary and TGIRmax, and early and late TGIR50% as secondary endpoints. For long acting insulins Cmax and Tmax might become meaningless because of the flat time action profile of these kinds of insulin formulations.
Clinical safety studies should be performed to ensure a comparable adverse event profile, especially with regard to hypoglycaemia and local skin reactions. Immunogenicity should be evaluated by the measuring the incidence and titres of antibodies in studies with a reasonable number of patients with type 1 diabetes mellitus and an exposition of at least 12 months. After approval of the drug a pharmacovigilance system needs to be implemented to identify potential rare adverse events.