Aims/hypothesis: Recent epidemiological studies suggest that treatment with insulin glargine (A21Gly,B31Arg,B32Arg human insulin) may promote cancer growth. The present meta-analysis was performed to assess the risk of cancer during treatment with insulin detemir (B29Lys(epsilon-tetradecanoyl),desB30 human insulin), another long-acting insulin analogue.
Methods: This meta-analysis was performed in a population of 8,693 patients with type 1 or type 2 diabetes, who were included in Novo Nordisk-sponsored, randomised and controlled diabetes trials of at least 12 weeks in duration that compared insulin detemir with NPH insulin or insulin glargine. In a blinded manner, the adverse events with suspected treatment-emergent malignant tumours were obtained from these studies under three system-organ classes: 'Neoplasms benign, malignant and unspecified (including cysts and polyps)', 'Neoplasm' and 'Surgical and medical procedures'. Conditional ORs were estimated applying both the Mantel-Haenzel and Peto methods to ensure robustness of results.
Results: Separate analyses were performed for trials comparing insulin detemir with NPH insulin and insulin detemir with insulin glargine. In the first analysis, 16 studies were included with a total of 3,983 patients treated with insulin detemir and 2,661 patients treated with NPH insulin. In the second analysis, five studies were included with a total of 1,219 patients treated with insulin detemir and 830 patients treated with insulin glargine. The estimated OR for a cancer diagnosis between NPH insulin and insulin detemir was statistically significantly >1, with the ratio favouring insulin detemir. There was a more than twofold higher cancer occurrence in the NPH insulin-treated population. For the insulin detemir comparison with insulin glargine, there was a non-significant difference in ORs in favour of insulin detemir.
Conclusions/interpretation: In these randomised controlled diabetes trials, patients treated with insulin detemir had a lower or similar occurrence of a cancer diagnosis compared with patients treated with NPH insulin or insulin glargine, respectively.