A-333 Unusually Low Glycated Albumin Results Suggest Unexpected Interference in Study Samples from Pregnant Women

Abstract

Abstract Background Glycated albumin (GA) is an alternate marker of glucose control that is recommended in situations when hemoglobin A1c is unreliable. One of us (JPC) has examined its use in pregnant women with diabetes since 2019 and regularly observed GA values of 8.0–19.4%, with occasional outliers. However, in a recently recruited cohort of 149 women with and without gestational diabetes, 58% of the samples showed GA results <8.0%, despite no apparent physiological differences. Here we further investigate a subset of these samples using a liquid chromatography-mass spectrometry method (LC-MS/MS). Methods Glycated serum proteins (GSP) were measured using the GlycoGap kit (Diazyme Labs, Poway, CA), an enzymatic method, and subsequently converted to percent glycated albumin (GA%) using the equation provided by the manufacturer.The LC-MS/MS assay was performed according to a previously published protocol. The samples were first reduced and alkylated, then spiked with a blend of two isotope-labeled standards and digested using Glu-C enzyme. The resulting digests were analyzed using the QTRAP 6500+ (AB Sciex, Foster City, CA) coupled with a LC system (Shimadzu, Japan) in MRM mode to provide values representing the percentage of glycated to non-glycated Lys-525 in albumin (MSGA%).Samples from a previously collected cohort (A, n = 78) of pregnant women with and without diabetes, with GA% range of 8.5–24.2%, were analyzed using both methods, and a standard curve (MSGA% vs GA%) was established. Thirty-eight samples with GA <8.0% (cohort B) were then analyzed using the LC-MS/MS method, and the standard curve was used to generate the expected enzymatic GA% results. Results The comparison of the LC-MS/MS and Diazyme methods resulted in a straight-line equation (r2 = 0.84, slope = 1.81, y-intercept = 0.79). For the cohort B samples analyzed by LC-MS/MS, the median MSGA% result was 5.89% (range: 4.32–8.95%). Based on the equation derived from the method comparison (MSGA% vs GA%), the expected median value with the Diazyme enzymatic method would have been 11.48% (range: 8.63–17.02%). These GA% values are well within the typically expected range. Conclusion LC-MS/MS results and GA% calculated from a method comparison curve were within the expected range while original enzymatic results were not, suggesting the presence of an interfering substance. Our hypothesis is that prenatal supplements may lead to elevated ascorbic acid, which is known to cause interference in certain enzymatic assays. We are continuing to investigate this possibility.