Development and validation of a method for quantification of human insulin and its synthetic analogues in plasma and post-mortem sera by LC-MS/HRMS

Analysis of human insulin and its synthetic analogues is increasingly requested for clinical monitoring, for anti-doping purposes, but also for forensic cases. Indeed, insulin analogues may be abused for suicide or homicide - whence their forensic interest. Collection and storage conditions, as well as the phenomenon of degradation make post-mortem serum samples analytically challenging and consequently, the rate of exogenous insulin administration as cause of death is undoubtedly underestimated. However, with recent technological advances and the development of new extraction techniques particularly for anti-doping analyses, detection of insulins in post-mortem samples seems to be achievable. This study describes the first validated quantitative method for analysis human insulin and its six analogues (lispro, aspart, glulisine, glargine, detemir and degludec) in plasma and post-mortem sera. Various extraction processes, namely precipitation + solid phase extraction (SPE), filtration + SPE, precipitation + SPE + immunopurification, and filtration + immunopurification, were assessed to evaluate the lowest limit of detection for all target analogues. The selected sample preparation consists of filtration step followed by immunopurification extraction with an anti-body precoated ELISA plate for plasma. For post-mortem sera, the first step of precipitation was added to remove matrix interferences. The extracts were analyzed by ultra-high-performance liquid chromatography-high resolution mass spectrometry (LC-HRMS), interfaced by electrospray (ESI). The method was validated with respect linearity, precision, accuracy, recovery, matrix effect, dilution and carryover. The limit of quantification (LOQ) in plasma was 0.5 ng/mL for human insulin and rapid-acting insulins, 1.0 ng/mL for glargine, 2.5 ng/mL for degludec and 10 ng/mL for detemir. Two types of post-mortem sera were studied based on the post-mortem interval (PMI): inferior or superior to 48 h. The obtained LOQ were the same for each analogue, independent from the PMI: 1.0 ng/mL for human insulin and rapid-acting insulins, 1.0 ng/mL for glargine, 2.5 ng/mL for degludec and 10 ng/mL for detemir. At the LOQ level, for all insulins and all samples, accuracy was between 70 and 130% and precision inferior to 30%. The validated method was applied to five subjects participating in therapeutic monitoring of insulin and to seven post-mortem cases.