The term insulin analogues refers to a group of molecules that very closely resemble human insulin in amino-acid composition, structure and biological action. Semantically, the word insulin analogue implies that animal insulins such as porcine insulin (which differs from human insulin by only one amino-acid) and bovine insulin (three different amino-acids) can be considered insulin analogues as well, but generally the term is reserved for the custom-made insulins that have been biosynthetically engineered for specific pharmacokinetic purposes from the 1980's onwards.
Broadly, the current insulin analogues fall into two categories:
- the short-acting insulin analogues that were designed to work faster than human regular insulin, and
- the long-acting insulin analogues that were designed to work longer than the conventional NPH and Zinc insulins
Originally, one of the goals of developing long-acting insulin analogues was also to find an insulin with selectively increased action in the liver, because the major flaw in s.c. insulin therapy is that hepatic insulin levels are low and systemic insulin levels are high in comparison with the physiological situation (where insulin is directly released into the portal vein). This goal, however, has proven to be elusive and in retrospect seems unreasonable given the fact that the insulin receptor has no specific subtypes with a differential organ distribution. Most reports of hepatoselectivity of insulin analogues these days can not stand clinical scrutiny. It should be realized that at low insulin levels insulin mainly suppresses lipolysis and hepatic gluconeogenesis, and that higher insulin levels are needed to stimulate peripheral glucose uptake. Thus, any insulin that results in low insulin levels for a prolonged period of time (i.e. any true basal insulin) will seem more 'hepatoselective', but this does not imply a more adequate distribution of the s.c. injected insulin over the various organs.
Evidence for the use of insulin analogues
When following the strict criteria of the Cochrane Collaboration, remarkably few studies qualify as evidence for the use of insulin analogues over conventional insulins, and their meta-analyses can be summarized as follows:
- The use of short-acting insulin analogues in type 1 diabetes significantly but marginally reduces HbA1c by about 0,2% with a non-significant and marginal decrease in hypoglycaemia in comparison with regular human insulin. The benefits do seem to be a bit larger when using short-acting insulin analogues as substitute for human insulin in insulin pump therapy.
- The use of short-acting insulin analogues in type 2 diabetes neither reduces HbA1c nor does it decrease hypoglycaemia in comparison with regular human insulin.
- The use of the long-acting insulin analogues in type 1 diabetes in comparison with NPH insulin has been insufficiently studied
- The use of the long-acting insulin analogues in type 2 diabetes in comparison with NPH insulin has no demonstrable effect on HbA1c. While glargine seems to reduce hypoglycaemia rates by 16% (in absolute terms about 1 episode less per year!), no significant difference in hypoglycaemia rates has been demonstrated for detemir.
When taking into account other meta-analyses (which are usually less rigorous and/or less free from conflicts of interest) the overall picture is a bit better, but the analogues have not brought about the revolution that was hoped for when they were developed. Thus, for each individual patient the potential benefits of insulin analogue therapy have to be carefully weighed against the low cost, good efficacy and long-term safety of the conventional insulins.