Evaluation of a one-step sample preparation protocol for analysis of total serum testosterone by LC–MS/MS

Abstract

In this study, a one-step method for liquid-liquid extraction has been compared against a two-step procedure for testosterone assays in terms of accuracy, specificity, recovery, lipid removal and baseline noise, using QCs and unknown samples. The difference in accuracy was less than 5% for adult sera, while it was less than 10% for prepubescent sera. To compare specificity, the ion ratio transition of 289 → 97 to 289 → 109 was monitored for all QCs and unknown samples; no interference in the testosterone peak was observed for any tested sample prepared by either the one-step or two-step procedure. The baseline comparison of LC–MS/MS chromatograms of samples indicated that samples prepared by the one-step procedure were of the same quality as those prepared by the two-step procedure; however, recovery in unaltered serum using the one-step procedure was approximately 15% greater across the low to high concentration range. Furthermore, recovery using the one-step procedure was more consistent between stripped and unstripped serum. Lipids were removed efficiently in the two-step procedure as verified by monitoring the typical phospholipid MRM transition of m/z 496 → 184. For every sample processed by the one-step procedure, a one-minute online column wash with 95% methanol was able to remove 95% of the bound lipids, thereby providing a column life-time approximately equivalent to that for the two-step procedure. The presented data indicate that the one-step procedure could replace the two-step procedure while maintaining accuracy, saving time, increasing recovery, and minimizing the potential for errors with the fewer steps required.