Category: Professional Posters
Purpose: Insulin analogues were designed to have a more desirable pharmacokinetic/pharmacodynamic profile for insulin therapy, they have demonstrated with lower rates of hypoglycemia incidents in DM patients. In Taiwan, another reason why insulin analogues are widely used in diabetic patients is partly due to they are the only available pen-filled injections and are more patient-friendly compared to human insulin, which are only available in vial packaging. However, there has been a long debate on whether insulin analogues increase risk of cancer. This study aims to investigate the effect of insulin analogues on the risk of cancer development compared with human insulin.
Methods: This study is conducted using the Longitudinal Health Insurance Database 2010 (LHID2010), which contains one million individuals randomly sampled from the Registry for Beneficiaries of the National Health Institute (NHI) program in 2010 in Taiwan. The eligible study population for this retrospective cohort study is defined as patients with newly diagnosed diabetes (ICD-9-CM code 250.XX) between 2001-2011, and with a first-time prescription of any type of insulin. Patients aged under 20 and over 80 when diagnosed with diabetes, diagnosed with any type of cancer prior to insulin use, exposed with both human insulin and insulin analogues were excluded. The follow up period is defined as the date of first insulin prescription to the diagnosis of any cancer, death or the end of the database period (2011). Binary logistic regression analysis was conducted, and odds ratio (OR) and 95% confidence interval (CI) were estimated. Multiple variables were assessed, including sex, age at diagnosis of DM, type of DM, social economic status, comorbidity, baseline concomitant medication use, and number of hospitalizations during the follow up period.
Results: 2708 newly diagnosed DM patients with first-time insulin treatment were enrolled, 1252 used human insulin and 1456 used insulin analogues. The median follow up period were 6.15 and 6.76 years for human insulin and insulin analogue groups. 369 patients were diagnosed of cancer during the follow up period. The adjusted odds ratio for developing cancer in association with insulin analogues compared to human insulin was 0.575 (p<0.001). Subgroup analysis was conducted and demonstrated variables such as sex, type of diabetes, age at diagnosis of diabetes, diabetes disease duration, and baseline concomitant use of metformin and statins were effect modifying factors.
Conclusion: Insulin analogues demonstrated with reduced risk of cancer development compared to human insulin in DM patients, while the protective effect was lost in patients with disease duration longer then 9 years. This study showed decrease of cancer risk for insulin analogues, however, with only 369 cancer events observed, the clinical significance of this result should be determined in larger, long-term studies.