|
Abstract INTRODUCTION:
The simultaneous quantification of 10 hypoglycemic drugs via LC-MS/MS presents a "Goldilocks" challenge: balancing the sensitivity required for trace detection against the high concentrations encountered at the Upper Limit of Quantitation (ULOQ). Two critical factors often disrupt this equilibrium. First, the contribution of the Internal Standard (IS) to the negative control (blank) channel which can artificially inflate the baseline via isotopic impurities thus compromising the performance at the Lower Limit of Quantitation (LLOQ). Conversely, high-concentration samples introduce the risk of analyte contribution to the IS channel—either through isotopic overlap or "crosstalk”, leading to inaccurate normalization and inaccuracy at the ULOQ. When combined within a single assay, these opposing interferences create a narrow and often elusive analytical window.
OBJECTIVE:
The objective of this study was to develop an LC-MS/MS method for the simultaneous measurement of first-generation sulfonylureas (chlorpropamide, tolbutamide, tolazamide), second-generation sulfonylureas (glipizide, glimepiride, glyburide), thiazolidinediones (rosiglitazone, pioglitazone), and meglitinides (nateglinide, repaglinide) in human serum/plasma using a SCIEX Triple Quad™ 6500+ mass spectrometer. The study explores the optimization of sample preparation and mass spectrometry parameters to find the "just right" window: ensuring IS levels are high enough to provide robust normalization but low enough to avoid blank contamination, while simultaneously mitigating analyte-driven interference in the IS signal.
METHODS:
100 µL matrix matched calibrators and quality control samples were fortified with 10 µL internal standard solution (mixture of d4-chlorpropamide, d9-tolbutamide, d7-tolazamide, d11-glipizide, d4-glimiperide, d3-glyburide, d5-nateglinide, d4-pioglitazone, d4-rosiglitazone and d5-repaglinide) and evaluated using two extraction approaches: protein precipitation using acetonitrile and supported liquid extraction (SLE) using methyl tert-butyl ether (MTBE). Following extraction, samples were dried and reconstituted under initial chromatographic conditions. Separation was performed on a Phenomenex Luna 3 µm C8 (50 x 2 mm) column with gradient elution using 2 mM ammonium acetate and methanol as mobile phases A and B respectively. Analytical measurement ranges were 10-10,000 ng/mL for the first-generation sulfonylureas and thiazolidinediones and 1-1000 ng/mL for second-generation sulfonylureas and meglitinides. To address isotopic interference, multiple strategies were explored, including IS concentration balancing, use of internal standards with additional labeling, ionization mode switching, sample preparation, and adjustment of LLOQ.
RESULTS:
Significant isotopic crosstalk was observed for glimepiride, glyburide nateglinide, and repaglinide leading to non-linearity in calibration curves. When the IS concentration of these analytes was reduced from 50 ng/mL to 25 ng/mL, interference in blank sample reduced from 77% to 51% for glimepiride, 55% to 45% for glyburide, 135% to 45% for nateglinide, and 51% to 42% for repaglinide. Further reduction in IS concentration to 10 ng/mL shifted the interference levels to 40% (glimepiride), 48% (glyburide), 31% (nateglinide) and 30% (repaglinide). Switching the ionization mode from Electrospray Ionization (ESI) to Atmospheric Chemical Ionization (APCI) combined with substituting lower mass isotope analogues (d3-glimepiride and d5-glyburide) for higher mass analogues (d5-glimepiride and d11-glyburide) reduced the IS contribution for glimepiride to 0% and 16% for glyburide. Despite these changes, the contribution to LLOQ remained high for nateglinide and repaglinide. Change in sample preparation from protein precipitation to SLE provided limited benefit, with SLE offering modest reductions primarily for repaglinide (4%). In contrast, continuing to use IS concentration of 10 ng/mL and transitioning data collection from multiple individual experiments (high number of data points per peak) to scheduled MRM (sMRM) mode, presented a different impact for glyburide and nateglinide (20 and 23% respectively) whereas repaglinide and glimepiride showed negligible isotopic cross talk (<5%). Implementing m+2 isotope peak for analytes provided moderate benefit for glyburide. Ultimately, raising the LLOQ from 1 to 5 ng/mL, was required to consistently reduce interference below 20% for all the affected analytes. At the ULOQ, analyte-to-IS crosstalk was most pronounced for glyburide where interference exceeded 200%. This was mitigated to <5% by switching to a higher mass (d11) analogue, which provided a better mass buffer against the analyte’s natural isotopic distribution.
CONCLUSION:
The development of this multi-analyte assay was primarily hindered by significant isotopic interference, likely arising from either in-source isotopic scrambling or presence of unlabeled analyte impurities in IS stock solutions. Despite employing higher-mass isotopes and alternative extraction techniques like LLE, the crosstalk remained a persistent barrier to low-level quantification for some analytes. Notably, increasing the dwell time between measurements inadvertently raised crosstalk percentages in previously unaffected analytes. To achieve acceptable accuracy and precision, IS concentrations were reduced, higher-mass analogs were used for problematic internal standards, transitioned sample preparation from protein precipitation to SLE and the lower limit of quantification (LLOQ) was raised for specific compounds. These refinements successfully maintained interference below the 20% threshold required for validated bioanalytical methods. |
= Discovery stage. (57.21%, 2026)
= Translation stage. (23.38%, 2026)
= Clinically available. (19.40%, 2026)
