Cochrane Review – long-acting analogues vs NPH insulin in Type 2 diabetes

Cochrane Review – long-acting analogues vs NPH insulin in Type 2 diabetes

A Cochrane Review April 2007
K Horvath, K Jeitler, A Berghold, SH Ebrahim, TW Gratzer, J Plank, T Kaiser, TR Pieber, A Siebenhofer Cochrane Database of Systematic Reviews 2007 Issue 2 (Status: New)

 Insulin analogues are the latest form of GM synthetic insulin and this review compares long-acting analogues glargine [Lantus] and determir [Levemir] with long-acting ‘human’ isophane [NPH] insulin for Type 2 diabetes. For us to have an informed choice of treatment, it is necessary to look at evidence from high quality systematic reviews and Cochrane Reviews provide just such evidence.

The authors’ conclusions are:

"If at all there is only a minor clinical benefit of treatment with long-acting insulin analogues for patients with diabetes mellitus type 2 treated with "basal" insulin regarding symptomatic nocturnal hypoglycaemic events. Until long-term efficacy and safety data are available, we suggest a cautious approach to therapy with insulin glargine or detemir."

Below is the ‘Plain Language Summary’ but the full review can be found on the Cochrane Database

Plain language summary
No unambiguous clinical benefits of treatment with long acting insulin analogues in the majority of people with type 2 diabetes mellitus demonstrated NPH (Neutral Protamine Hagedorn) insulin is the current standard for basal insulin in the blood glucose lowering therapy in people with type 2 diabetes mellitus. The mode of action of this insulin is highly variable, which may be the cause for the difficulties some people with diabetes have to achieve current goals for long-term metabolic control. Therefore, new insulins which are thought to show more favourable properties of action have been developed: insulin glargine and insulin detemir. Because of their theoretical advantages, it is thought that treatment with these new insulin analogues might lead to a beneficial effect, for example less hypoglycaemia or a better metabolic control, possibly resulting in higher quality of life and treatment satisfaction less late diabetic complications such as problems with eyes, kidneys or feet and myocardial infarction, stroke or death.

Although epidemiological studies indicate that high concentrations of blood glucose carry a higher risk for these late complications, evidence for a beneficial effect of glucose-lowering therapy is conflicting. Following from the different results of large clinical trials, interventions seem to carry different substance specific beneficial or adverse effects. As a consequence, conclusions on the effects of different blood glucose lowering interventions on these outcomes can not be drawn from their effect on blood glucose concentration alone.
Methodological quality of all the studies was rated low ("C"). Eight studies investigated altogether 2293 people. Trials lasted between 24 and 52 weeks. Our analysis of the currently available long-term trials comparing long acting insulin analogues with NPH insulin showed that insulin glargine and insulin detemir were almost identically effective compared to NPH insulin in long-term metabolic control (HbA1c). Fewer people experienced symptomatic overall or nocturnal hypoglycaemic episodes with treatment with either of the two analogues. No conclusive information on late complications or on possible differences in the number of fatalities exists. For insulin glargine one study found a higher rate of progession of diabetic retinopathy in patients treated with insulin glargine, while in another investigation the opposite result was found. It was thus not possible to conclude for certain whether insulin glargine treatment is safe or not.

From the retrieved trials it was also not possible to draw firm conclusions on the effects of these new insulins on quality of life or their cost effectiveness. Until long-term data on benefit and risk are available, we suggest a cautious approach to treatment with insulin glargine or insulin detemir.

If this language is not plain enough, let’s make it plainer…

  • We know that Lantus, Levemir and human long-acting insulins are the same in terms of blood glucose control as measured by HbA1s.
  • We know that fewer people in the studies experienced symptomatic overall or night hypos with both the two analogues but we don’t know about the numbers of hypos without warnings [asymptomatic].
  • We don’t know if treatment with Lantus and Levemir results in more or less complications over time or if there are any differences in death rates.
  • We don’t know if Lantus causes higher rates of retinopathy – one study showed it did and one that it didn’t, so we don’t know if it’s safe or not.
  • We don’t know if these insulins improve quality of life or not.
  • We don’t know if they are cost-effective or not.
  • We do know that the authors recommend a cautious approach to their use.

From this review we know two things – there are an awful lot of uncertainties about long-acting analogues and the authors’ recommendation for a cautious approach to prescribing these insulins is not being adopted in the UK and many other countries.