Datamonitor attended this year's European Association for the Study of Diabetes meeting in Vienna, Austria. Following the issuance of guidance on a potential link between Lantus and increased cancer rates in June 2009, presenters at the conference discussed the current evidence and its limitations. However, a potential mechanism by which insulin could promote cancer cell growth was also proposed.
Although the European Association for the Study of Diabetes (EASD) recommended the continued use of Lantus in type 1 and type 2 diabetics, it has called for increased research into the relationship between insulin therapy - particularly Lantus (insulin glargine; Sanofi-Aventis) - and the risk of cancer.
It was postulated during the discussion at the 2009 EASD meeting that insulin use may be related to the increased incidence of cancer seen in diabetics over that seen in the general population. The preliminary data discussed in the symposium suggested lower rates of cancer with Levemir (insulin detemir; Novo Nordisk) compared to neutral protamine Hagedorn (NPH), but higher rates with Lantus.
Since the EASD announcement, there has been little change in the use of Lantus due to the weakness of the four insurance studies published in June 2009, but the presentation of data at the EASD to such a large audience of specialists may lead to a change in prescription patterns. It is unlikely that existing patients will be switched due to the biovariability differences between Lantus and Levemir and the loss of glycemic control during titration, but physicians may think twice when initiating a patient on a long-acting insulin. Long-term implications of the cancer scare may involve a greater focus on reducing the insulin requirements of patients through reductions in beta-cell function decline and insulin sensitivity, rather than using insulin as a panacea to elevated glycated hemoglobin A1c (HbA1c) levels.
Dr Jay Skyler of the University of Miami, presenting a defense of Lantus on behalf of Sanofi-Aventis, highlighted the caution taken by the EASD when analyzing the data and making its recommendation that patients continue using Lantus. However, he was critical of the role of the lay-media in sensationalizing the story without reporting the caveats included within the original EASD press release. Dr Skyler also quoted from a rebuttal of the dose adjustments used within the studies which stated that the technique was flawed and the potential link between Lantus and cancer was unwarranted.
Analysis of data from a clinical trial to determine the impact of Lantus therapy on rates of diabetic retinopathy found no increase in malignancy and the medical ethics committee of the ORIGIN clinical trial found no reason to discontinue Lantus use in the trial due to increased malignancy rates. Dr Skyler concluded that Lantus was safe to use and that there was no link between the drug and cancer. Prior to the EASD meeting, however, Sanofi-Aventis announced that it will be undertaking a retrospective cohort study in Northern Europe and an epidemiological study in the US to determine the safety of Lantus.
Following the defense of Lantus, Dr David Russell-Jones from the University of Surrey, UK, offered a defense of the other long-acting insulin analog on the market, Levemir. Having not been included in the original insurance database studies, it was unknown whether Levemir would demonstrate similar characteristic to Lantus. A meta-analysis of the Novo Nordisk clinical trials database demonstrated a cancer event rate of 0.36 events/100 years exposure with Levemir compared to 0.92 events/100 years exposure with NPH. Data from a meta-analysis of Levemir and Lantus produced event rates of 0.87/100 years and 1.27/100 years exposure, respectively. Both of these results, however, were insignificant.
Dr Russell-Jones then focused on preclinical studies of the insulin analogs, long-acting as well as short-acting, which had been conducted by Novo Nordisk. These studies demonstrated that Lantus bound with significantly greater affinity to the IGF-1 receptor than Levemir, NovoRapid (insulin aspart; Novo Nordisk) and Humalog (insulin lispro; Eli Lilly), and that Levemir had lower mitogenic impact on selected cell lines than human insulin.
An initial study of 127,000 patients in an insurance database in Germany raised the possibility of a greater incidence of cancer in patients receiving Lantus compared to traditional human insulin NPH. This was submitted to Diabetologia, the journal of the EASD, but the editors treated the paper with understandable caution, wary of causing an unnecessary safety scare. Consequently, the editors delayed the publication of the paper and asked for confirmation of the study through similar studies of databases in Sweden, Scotland and the UK. The Swedish study found a statistical link between Lantus and breast cancer while the Scottish study identified a non-statistical link with breast cancer. In the UK study, meanwhile, no link was found. Importantly, the four studies only looked at the safety of Lantus compared to NPH with no other insulin analogs investigated.
A number of criticisms and weaknesses of the studies have been highlighted, primarily the make up of the patient populations. After adjusting for a number of variables, the patient populations in the Lantus and NPH groups were different and the pre-treatment characteristics of the groups may result in the differences seen in cancer risk. Additionally, the low number of cancer events in the Swedish and Scottish studies could mean that they were a result of chance.
During the symposium, Dr Ulf Smith, chairman of the EASD, and Dr Edwin Gale, editor of Diabetologia, outlined a potential mechanistic pathway for insulin to have mitogenic properties. Throughout the session it was emphasized that insulin does not have oncogenic (cancer inducing) properties, but could potentially act to promote the growth, or mitogenesis, of cancer cells through the IGF-1 receptor.