MSACL 2017 US Abstract

Native Matrix/Surrogate Analyte Calibration: Quantifying Testosterone with Deuterated Testosterone Calibrators

Joshua Hayden (Presenter)
Weill Cornell Medical College

Bio: Dr. Joshua Hayden is an Assistant Professor of Pathology and Laboratory Medicine at Weill Cornell Medical College. He serves as the director of the Toxicology and Therapeutic Drug Monitoring laboratory and assistant director of the Central Laboratory at New York Presbyterian Hospital. He is a board certified clinical chemist with a special interest in the development and implementation of mass spectrometry assays for clinical diagnostics. Prior to joining Weill Cornell, Dr. Hayden obtained a PhD in chemistry from Carnegie Mellon University and went on to conduct postdoctoral research at the Massachusetts Institute of Technology. He then completed a two year clinical chemistry fellowship at the University of Washington, where he trained under the mentorship of Dr. Geoffrey Baird and Dr. Andrew Hoofnagle.

Authorship: Chase Mazur (1), Rich Tyburski (2), Joshua Hayden (3)
(1) and (3) Weill Cornell Medical College, (2) Isosciences

Short Abstract Our objective was to develop a calibration approach that does not require analyte-free matrix. Towards this end, we validated a testosterone assay that utilizes deuterated testosterone in pooled patient serum as calibrators. The assay involved a simple liquid-liquid extraction and carbon-13 labeled testosterone as the internal standard. Testosterone values for patient samples were obtained with standard calibrators and native matrix/surrogate analyte calibrators (deuterated testosterone in pooled patient serum). Accuracy of the native matrix/surrogate analyte approach was confirmed by comparison with the traditional calibration approach (slope 1.04, R2=0.99) and an established LC-MS/MS assay (slope 1.07, R2=0.996). The success of this approach suggests that authentic matrix/surrogate analyte could be a useful approach for samples where analyte-free matrix is challenging to obtain.

Long Abstract

Introduction:

It is a commonly held misconception that results from liquid chromatography tandem mass spectrometry (LC-MS/MS) represent an unquestionable gold standard for small molecule quantitation. In truth, the accuracy of LC-MS/MS assays requires extensive assay validation and careful consideration for assay calibration. It is especially challenging to develop calibration strategies for endogenous analytes when analyte-free authentic matrix is hard to find. In these cases, either an analyte-free, surrogate matrix or a chemically treated (charcoal stripping) “authentic” matrix can be used. . One intriguing, though infrequently reported approach (1,2), is be to utilize calibrators prepared with authentic matrix and stable, isotopically labelled “surrogate” analyte (deuterated, carbon-13 etc). To our knowledge, this approach has never been used in a clinically validated LC-MS/MS assay.

Methods:

Two types of calibrators were prepared for use in the testosterone (Te) assay described below. The first involved spiking testosterone into a surrogate matrix (SigMatrix, MSG4000). The second involved adding testosterone-[16,16,17-d3] (d3-Te) as the surrogate calibration analyte at varied levels to pooled serum. The serum pool was prepared by mixing male serum (high Te pool) or pooled female serum (low Te pool). Te and labeled-Te were obtained from Isosciences. For testosterone analysis, patient samples or calibrators were mixed with testosterone-[2,3,4-13C3] as an internal standard. A liquid-liquid extraction was then performed with methyl tert-butyl ether (MTBE) at a 3:1 ratio. Samples were vortexed for 60s, left to sit for 30 minutes and then centrifuged at 10,000g for 4 minutes. The supernatant (150uL) was aspirated into a new tube and dried down at room temperature. The samples were concentrated 3X by reconstituting with 50uL of mobile phase starting conditions (30% methanol:water) and transferred to LC-MS autosampler vials. Separation and sample analysis was performed on an Agilent 1290 Infinity II LC coupled to an Agilent 6495 Triple Quad LC/MS with Agilent MassHunter software. The LC separation column used was an Agilent Poroshell 120 C18 (2.1x50mm, 2.7um) heated to 50C. Mobile phases used were A: 5mM ammonium formate in H2O and B: 5mM ammonium formate in methanol. The MRM transitions used: testosterone 289.2→97 and 289.2→108.8 m/z; d3-testosterone 292.1→97 and 292.1→109 m/z; 13C3-testosterone 292.2→100.1 and 292.2→112 m/z.

Results:

Using calibration curves built with d3-Te, the LLOQ for Te in this assay was 10 ng/dL (100 pg/mL). The analytical measurement range (AMR) for Te in the assay was 10 – 10,000 ng/dL (100 – 100,000 pg/mL). Comparative analysis of patient samples using this surrogate analyte/native matrix calibration approach with a more traditional authentic analyte/synthetic matrix approach showed good correlation (slope 1.04, R2=0.99). Comparison to an established clinical Te assay also showed acceptable agreement for patient results obtained with the surrogate analyte/native matrix calibration approach (slope 1.07, R2=0.996). No difference in results was seen if high or low testosterone serum pools were used for the authentic matrix/surrogate analyte calibrators. Use of a single patient’s serum for the calibrators did increase the error of the obtained results.

Conclusion:

We have demonstrated this model of calibration built using a surrogate analyte (d3-testosterone) spiked into native matrix (pooled serum) to be effective for the sensitive and accurate measurement of Te. Using our approach, we have bypassed the need to use synthetic or analyte-free matrices in order to mitigate endogenous interference. Going forward, this method could be applied to a number of additional analytes that require calibration in analyte-free or synthetic matrices.

References & Acknowledgements:

References:

1. Quantitation of Endogenous Analytes in Biofluid without a True Blank Matrix Wenlin Li and and Lucinda H. Cohen. Analytical Chemistry 2003 75 (21), 5854-5859

2. Mass Spectrometry–Based Candidate Reference Measurement Procedure for Quantification of Amyloid-β in Cerebrospinal Fluid. Andreas Leinenbach, Josef Pannee, Thomas Dülffer, Andreas Huber, Tobias Bittner, Ulf Andreasson, Johan Gobom, Henrik Zetterberg, Uwe Kobold, Erik Portelius, Kaj Blennow. Clinical Chemistry Jul 2014, 60 (7) 987-994

Acknowledgments:

We would like to thank Brian Rappold for thoughtful discussion.

Financial Disclosure

IP Royalty: no

Planning to mention or discuss specific products or technology of the company(ies) listed above:

yes