MSACL 2024 Abstract

INTRODUCTION: Highly accurate and precise quantitation of antibodies in the clinical setting, especially at levels below a few micrograms per milliliter, can be very challenging due to interferents. High-resolution mass spectrometry (MS) and state-of-the-art spectral deconvolution software offer major advantages to address this challenge.

OBJECTIVES: Vedolizumab light chain from human serum samples was quantitated using liquid chromatography-MS (LC-MS) and two different data analysis methods. Major objectives were to introduce Gábor Transform deconvolution software, which has never before been demonstrated for protein quantitation, and to compare light chain concentrations determined using this method to those determined using commercial software.

METHODS: Nivolumab standard is added to human serum samples containing vedolizumab before Melon Gel purification and cleavage to produce light chains. LC-MS data for quantitation were acquired for ~400 samples in parallel at Mayo Clinic and the University of Oregon, where they were analyzed using an extracted ion chromatogram-based method or Gábor-Transform-based software (“iFAMS Quant+”), respectively. These samples ranged in concentration from below 1 microgram/milliliter to ~200 micrograms/milliliter (~3 orders of magnitude). Concentrations determined using the two methods were plotted against one another to determine method bias and Pearson R2 coefficients, and Bland-Altman analysis was used to identify concentration-dependent differences in the results from the two methods.

RESULTS: Due to the high mass accuracy of the MS instrumentation used, signals for vedoluzimab and nivolumab light chain were easily distinguished from those of endogenous IgG light chains and other interferents on both instruments. Quantitation was highly consistent between the two instruments and software methods, with R2 coefficients for each sample set comparison between Mayo and U Oregon results exceeding 0.9. iFAMS Quant+ was able to establish a Lower Limit of Quantitation (LLOQ) of less than 2 microgram/milliliter for these samples. As expected, Bland-Altman analysis revealed that the largest discrepancies between the two methods typically occur at the lowest concentrations studied and are consistent with the LLOQ.

CONCLUSION: Highly chemically specific antibody drug quantitation can be achieved using LC-MS and state-of-the-art quantitation software. Excellent agreement between results for vedolizumab light chain in human serum at two different institutions with different instrumentation and data analysis software, especially at low-microgram per milliliter concentrations, highlights the advantages of this technique to address the challenge of antibody quantitation. These results also constitute the first demonstration of Gabor Transform deconvolution tools for protein quantitation.