LAUSR

Insulin use has dramatically increased over the past 30 years, in part because of the increasing realization of its role in the treatment of type 2 diabetes (T2D) (Figure 1A). Worldwide, the prevalence of type 1 diabetes (T1D) is 0.95 per thousand persons, for a total of approximately 7.5 million persons who require insulin in multiple daily dosages out of the world population of 7.9 billion. In addition, an estimated 30.2 million persons with T2D are treated with insulin globally, using more than 500 billion units of insulin per year, at a cost of some US$20 billion (Figure 1B). In the United States, 1.6 million persons have T1D of the total of 6.0 million people with diabetes who use insulin (Figure 2). Many people with diabetes use biosynthetic human insulin (BHI) produced by recombinant DNA technology. In 1996 the first rapid-acting insulin analog became available. This molecule, insulin lispro, is formulated by interchanging the positions of the amino acids lysine and proline at the end of the B chain of the insulin molecule, resulting in weaker self-association than BHI, with consequent greater rapidity of onset and shorter duration of action. Four years later, in 2020, the first long-acting analog was approved, insulin glargine, with addition of two arginine residues to the B-chain and substitution of one of the asparagine residues of the A-chain by glycine, allowing the molecule to be formulated as an acidic solution that is slowly and evenly absorbed. Although some consider human insulin preparations to be similar in efficacy and benefit to insulin analogs, there is reasonable evidence suggesting that insulin analogs are preferable. Compared with (eutral protamine Hagedorn (NPH), long-acting insulin analog treatment is associated with significant reduction in hypoglycemia, as well as with evidence of improvement in glycemic control. Postprandial hyperglycemia, nocturnal hypoglycemia, and HbA1c are reduced to a greater degree in people with T1D receiving insulin aspart and other rapid-acting analogs rather than BHI, and reduction in postprandial hyperglycemia is seen in people with T2D comparing these two prandial insulin preparations. There has therefore been great interest in approaches to increase availability of analog insulin preparations for people with diabetes. Part of the impetus to the development of such products has been the high cost of insulin analogs, with the average price of rapid-acting and long-acting insulin analogs particularly high in the United States, between 8and 20-fold that in other developed countries, costing nearly US$6000 per insulin-treated person annually, with estimates that widespread availability of biosimilar insulin analogs could reduce this cost more than 50-fold. As a recent commentator observed, “The WHO added longacting analogues to its Essential Medicines List in its 2021 revision... Some people undoubtedly benefit from their use... Cost apart, there is no very good reason not to use them.” The US Food and Drug Administration (FDA) has suggested three levels of similarity that can be used in characterizing complex biologic pharmaceuticals such as insulin, with the goal of helping prescribers to offer appropriate biosimilar insulin analogs to people with diabetes treated with existing more expensive products. The FDA defines a follow-on biologic as one sufficiently similar to the original FDA-approved biologic to permit reliance on existing scientific knowledge about safety/ effectiveness, noting that this is determined on a case-bycase basis. A biosimilar is a follow-on biologic that is “highly similar to the reference product notwithstanding minor differences in clinically inactive components,” and having “no clinically meaningful differences from reference product in safety, purity, and potency.” Finally, the FDA has proposed that a biosimilar be termed “interchangeable” if the prescriber can expect the same clinical result as with the reference product in any given patient, and if “the risk in terms of safety or diminished efficacy of alternating or switching between use of the [biosimilar] and the reference product is not greater than the risk of using the reference product without such alternation or switch.” The distinctions made between the three levels of similarity may be necessary for regulatory purposes but are clearly somewhat arbitrary. What is the difference between “sufficiently similar” and “highly similar”? In DOI: 10.1111/1753-0407.13267