Limitations Of Insulin Assays For Work Up Of Patients With Hypoglycemia
Presentation Number: SAT-096
Date of Presentation: March 17, 2018, 2018
Kornelia Galior, PhD, Jennifer Fahse, BS, Ravinder Jit Singh, PHD.
Mayo Clinic, Rochester, MN, USA.
Iatrogenic hypoglycemia is the major limiting factor in achieving optimal glycemic control in diabetic patients, which is the result of excessive exogenous insulin administration. It is 10-fold more prevalent in type 1 diabetes than in type 2 diabetes mellitus and can lead to frequent morbidity. Ascertaining the cause of hypoglycemia is a major challenge because many commercially available insulin immunoassays lack the sensitivity required to detect and differentiate therapeutic insulin administration. This consequently provides misleading results and prevents the correct monitoring of patient dosing. Herein, we compared detection of common therapeutic analogs using commercial immunoassays and mass spectrometry. Mass spectrometry analysis was able to differentiate synthetic analogs, while immunoassays could not determine their identities. As a result, we used mass spectrometry to identify the presence of therapeutic insulins in the serum of diabetic patients. Specifically, commercially available serum was spiked with five different analogs: short lasting analogs, Lispro, Aspart, and Glulisine and long lasting analogs, Detemir, and Glargine. Samples were analyzed using six immunoassays, including Beckman Access, Unicel DxI, Abbott Architect, ADVIA Centaur XP, Siemens Immulite 2000, and Roche Cobas e602 and was also evaluated on Qtrap MS/MS for comparison. Clinical validation was performed on serum samples collected from 49 diabetic patients receiving insulin analogs as insulin replacement. Samples were extracted and subsequently eluted on a Versette Automated Liquid Handler using Mass Spectrometric Immunoassay (MSIA) pipette tips. The identity and quantification of each exogenous analog were determined using QExactive Plus Orbitrap LC-MS/MS system. Results demonstrate that mass spectrometry is able to detect long lasting insulin analogs, Glargine and Detemir, with high accuracy, 78% and 100%, respectively. In addition, short lasting insulin analogs, Aspart, Lispro and Glulisine were detected with 56%, 62% and 100% accuracy, respectively. In contrast, Roche e602 immunoassay did not detect exogenous insulins while other immunoassays had high cross-reactivity with insulin analogs, therefore were not able to differentiate their correct type. In conclusion, we were able to identify and distinguish therapeutic analogs in patients’ serum using mass spectrometry, which may have clinical utility in differential diagnosis for the cause of hypoglycemia.
K. Galior: None. J. Fahse: None. R.J. Singh: None.