MSACL 2022 Abstract

Liquid chromatography-mass spectrometry analyses have established a critical role in clinical laboratories across the globe due to the high specificity and sensitivity of the technique. However, maintaining accuracy of measurements is a never-ending challenge as quantification performance varies between instruments, compounds, different samples, and within- and across runs. Therefore, normalization with analyte-similar internal standards (IS) and use of IS-corrected multipoint external calibration curves for each analyte is the industry standard, but this approach is laborious and resource intensive. To overcome this challenge, we have developed an internal calibration method that utilizes the natural isotope distribution of the selected internal calibration material (IsoC) to generate a multipoint calibration curve that eliminates the need for frequent external calibration.
To perform measurements in this manner, the IsoC is spiked into each individual sample in place of the traditional IS, the samples are extracted as appropriate, and analyzed via LC-MS/MS. Following LC-MS/MS analysis, an internal calibration curve is generated for each sample by multiplying the IsoC concentration by the theoretical relative abundance of each isotope to generate a series of x-coordinates that are plotted versus the isotope intensity (y-coordinate). The resulting linear regression equation is then applied to each isotope on the analyte and those results are averaged to yield a final concentration for the analyte in the sample.
We first applied this technique to insulin-like growth factor-1 (IGF-1) serum measurements using a Q Exactive Plus high-resolution mass spectrometer and compared the results to reference methods using external calibration curves. For IGF-1, we tested both a single amino acid variant (A70T) and 15N labeled IGF-1 as IsoCs. When using linear regression analysis to compare IsoC results to reference methods, we observed slopes of 0.91 and 0.88 and R2 values greater than 0.99. Interday %CVs were less than 20% when using the single amino acid variant IsoC and within 10% when using the 15N IsoC. Comparison to reference material was also very favorable, with results within 7% of expected at concentrations greater than 50 ng/mL.
We have also assessed the viability of the technique for analysis of cortisol and cortisone in urine, which was also performed using the Q Exactive Plus system. For these analytes, 13C and deuterium labeled molecules were tested as IsoCs and the results were once again compared to a reference method. The IsoC method performed well for cortisol, with slopes indicating agreement within 10% and R2 of greater than 0.98 when using linear regression to compare to the reference methods. Cortisone did not perform as well, with slopes indicating agreement within 20% and R2 as low as 0.88. Interday %CVs for these analyses were all under 20% and predominantly less than 10%.
As indicated by the results above, this technique is best suited for moderate size molecules such as IGF-1, which inherently produce larger isotope distributions allowing for more points in the internal calibration curve. However, these data also demonstrate that IsoC can still produce satisfactory results for smaller molecules such as cortisol while greatly reducing the burden on clinical laboratories by eliminating the need for preparation, maintenance, and storage of daily use multi-point calibration curves. We believe this technique has the potential to supplant the traditionally used and laborious techniques while still maintaining the highest level of data integrity.