Insulin and cancer

I recently received an e-mail from an organization to which I belong, with the worrisome title "'Possible link' between taking insulin glargine and developing cancer." It went on to explain that the current issue of Diabetologia, the journal of the European Association for the Study of Diabetes (EASD), features four studies that point to a "possible link" between taking insulin -- especially insulin glargine (Lantus) -- and developing cancer. Diabetologia's website has a listing of involved studies, editorial comments, information for patients, and a press release from the manufacturer.

The series of articles brings up all sorts of questions: Do other newer long-acting insulin analogs such as Levemir (insulin detemir) also share this association? If Lantus is associated with cancer, what kind of cancer? Does the administration of insulin cause a new cancer, or cause a pre-existing cancer to grow faster than if there had been no insulin therapy? None of these questions are answered in the studies that are published here.

The impetus for the claim of a link was a German observational study of 127,031 patients who were started on insulin therapy; the authors found a positive association between cancer incidence and insulin dose (bigger insulin doses = higher incidence of cancer) for all insulin types that they studied, which included human insulin and several insulin analogs (glargine/Lantus, lispro/Humalog, and aspart/Novolog/Novorapid). After rejiggling the numbers, they concluded that "the cancer incidence with glargine was higher than expected compared with human insulin." There were several interesting limitations to the study: patients simultaneously on both analog and human insulin were excluded; and most important, there was no mention of what cancers were found: were they typical of what might be expected in people with diabetes, or some unexpected tumor type?

The second study was in the UK, and involved 62,809 patients, came to the opposite conclusion about insulin analogs (such as Lantus): "Use of insulin analogues was not associated with increased cancer risk as compared with human insulin."

The third study was in Sweden, followed 114,841 patients on insulin, and concluded "No definitive conclusions regarding a possible causal relationship between insulin glargine use and the occurence [sic] of malignancies can be drawn from the results of this study."

The fourth study evaluated 49,106 patients in Scotland. The authors found that patients "receiving any insulin glargine ... had the same incidence rate for all cancers as those not receiving insulin glargine" and concluded that "insulin glargine use was not associated with an increased risk of all cancers or site-specific cancers."

There has always been a suspicion that insulin or insulin analogs might be causally associated with bad things. Since insulin is a "growth factor," cancer has been on the list. The possibility that insulin analogs could be associated with cancer was clearly indicated years ago, when one insulin analog (called B10Asp) was found to induce breast tumors in rats; further development of that analog was halted, and subsequent analogs have not had such an association.