MSACL 2024 Abstract

INTRODUCTION:
According to the free hormone hypothesis, only the unbound or free fraction of the hormone exerts biological effects. Diagnosis and monitoring of thyroid disorders often requires the measurement of free T4 (fT4) and free T3 (fT3). Most clinical laboratories measure these hormones with immunoassays. Significant biases observed in immunoassays have been avoided by utilizing the sensitivity and specificity of liquid chromatography tandem mass spectrometry (LC-MS/MS). While ultrafiltration and size exclusion chromatography can be valid methodologies to extract free hormones, they are subject to variability due to protein leakage, temperature control and nonequilibrium. Equilibrium dialysis (ED) is considered to be the gold standard method for separation of free hormones from biological samples. Accuracy and specificity of detection are also of prime importance, considering very low endogenous concentrations of free hormones.

OBJECTIVES:
Our primary objective was to reduce the complexity and enhance sensitivity of an existing LC-MS/MS fT4 method that utilizes direct injection of the serum dialysates with 2-dimensional (2D) chromatographic separation (1st dimension separation being online SPE), and detection on a Triple QuadTM 5500 (Sciex, MA).

METHODS:
200 µL aliquots of serum samples and controls were subjected to equilibrium dialysis for 20 ± 1 hours. 150 µL of calibrators, and dialysates were mixed with 150 µL of internal standard 13C6 T4. Desalting of the dialysates was carried out on a Strata-X-AW 33 µm polymeric weak anion exchange, 10 mg (Phenomenex, CA) SPE plate. The sorbent was conditioned with methanol, equilibrated with water, followed by loading of sample. Two 200 µL aliquots of 10 % NH4OH in methanol were used to elute the analyte of interest. After evaporation under a flow of nitrogen at 60oC, the samples were reconstituted with 125 µL of 30% methanol in water and 50 µL was injected for analysis. Separation was performed on a 2.1 X 50 mm, 2.6 µm Kinetex C18 column (Phenomenex, CA). Mobile phase A was 0.05% formic acid in water and mobile phase B was a mixture of 1:1 Methanol: Acetonitrile. LC gradient consisted of 30% mobile phase B for 2.5 min, followed by 1 min conditioning with 95% mobile phase B and re-equilibration to the initial conditions. T4 peak eluted at 2.7 min with a total run time of 5.5 min. Sample analysis was performed on a Triple QuadTM 7500 (Sciex, MA). Data acquisition was performed using electrospray ionization in positive mode, using mass transitions of m/z 777.7>731.7 (T4 quant), 777.7>604.7 (T4 qual) and 783.7>737.7 (13C6-IS T4 quant) and 783.7>610.7 (13C6-IS T4 qual). Data analysis and quantitation were performed on OS software (SCIEX, OS 2.2.0).

RESULTS:
Lower and upper limits of quantitation (LLOQ and ULOQ) were 0.1 and 10 ng/dL respectively. Intra- and inter-day precision evaluated using three matrix-matched quality control samples (0.14, 2.05, 10.0 ng/dL) showed < 10% imprecision. Post-dialysis spiked samples spiked with T4, showed recoveries ranging between 85% and 115%. The method showed no carryover in a sequence of low-high-low T4 injected samples. Good agreement was observed with the existing LC-MS/MS method utilizing 2D LC separation for fT4 on the Triple QuadTM 5500 (slope = 0.997, r = 0.963, n = 53). Chromatograms of calibrators and patient samples (fT4 concentration ~0.11 ng/dL) demonstrated 7-10x improvement in peak area and signal-to-noise ratio, as compared to the existing method.

CONCLUSION:
We reduced the complexity of chromatographic separation by transitioning the method from 2D to 1D HPLC separation. Incorporating SPE based desalting of the dialysates and the improved detector sensitivity of the SCIEX Triple QuadTM 7500, allowed to reduce the lower limit of quantitation. Reducing injection volume from 250 µL to 50 µL permitted sufficient volume for additional injections if necessary for the reanalysis of samples (which is useful for pediatric sample testing). The new method provides significant enhancement in the signal intensity, and signal-to-noise ratio.