Difficulties in interpreting a therapeutic blood concentration of insulin lispro in postmortem blood: A case report

Abstract

Illustrate the complexity of interpreting a postmortem therapeutic blood concentration of insulin in a probable case of suicide using insulin lispro. The identification and quantification of insulin and its analogues continues to be a challenge in the field of forensic toxicology. Its use is not always limited to the treatment of diabetes but is often used illicitly. Numerous cases of suicides and suicide attempts have been described in the literature and nevertheless, this number still seems to be underestimated due to both the inability of laboratories to identify and discriminate between synthetic analogues of human insulin and due to the chemical instability of insulin especially in postmortem samples. Our case concerns a 63-year-old man who was found dead by his brother at home. Three empty Humalog® insulin pens and a letter mentioning his suicide intention were found next to his bed. At the request of the prosecutor, an external examination of the corpse was performed and revealed only an asphyxial syndrome not specific to a cause of death. The presence of a blood glucose monitoring device on the right arm was also observed, suggesting that the subject was diabetic. External examination was unable to determine the exact cause of death. For this reason, the pathologist was oriented on a toxic death. In order to proceed to further toxicological investigations, peripheral blood, urine and hair samples were collected. It is important to underline that, due to the long period elapsed between the external examination and the pathologist's conclusion (enabling the initiation of the toxicological analysis) and due to administrative delays, insulin testing was performed only 9 months after the samples collection. Blood and urine samples were stored at +4 °C until analysis. The identification and quantification of insulin lispro was performed in the postmortem blood after precipitation in the presence of bovine insulin (internal standard), by a mixture of acetonitrile/methanol (ratio 1:1) followed by a step of ultrafiltration using Amicon® centrifugal filters (3 kDa) and a step of immunopurification using Iso-Insulin ELISA coated wells. The analysis was performed with a UHPLC- Q-TOF (XEVO G2-XS). Detection of insulin lispro was performed using the precursor molecular ion m/z 1162.543 (charge state 5 +). Since human insulin and insulin lispro share the exact same molecular mass (5807 Da), discrimination cannot be made using the multi-charged molecular ion. The distinction is based on fragment m/z 217 which is present only in insulin lispro spectrum. Toxicological tests performed on the blood sample did not reveal any xenobiotic of toxic nature. Insulin lispro was detected in the femoral blood sample at 1,1 ng/mL, whereas human insulin was not detected. Following the administration of a therapeutic dose of 0.2 U/kg of insulin Humalog ®, a maximum concentration of 1.66 ± 0.42 ng/mL is achieved. Consequently, the concentration found in our case was therapeutic. Nevertheless, given the unstable nature of insulin, especially in postmortem blood, it is not possible to exclude that death was caused by an insulin overdose. The long period between collection and analysis (9 months), the hemolyzed state of the blood, the storage of the sample in unsuitable tubes and the fact that we were able to detect even small concentrations of insulin lispro suggest that the concentration at the time of death was much higher. The pathologist concluded that the cause of death was probably an overdose of insulin lispro. This case highlights that the interpretation of a postmortem insulin concentration requires a complex approach involving the assessment of many factors such as the subject's medical history and the time, mode and state of preservation of the sample.