Abstract
Background
Testosterone measurement by liquid chromatography tandem mass spectrometry (LC-MS/MS) is well accepted as the preferred technique for the analysis of testosterone. Variation is seen between assays and is likely to be due to method differences. One area of inconsistency among assays is the choice of internal standard. We investigated the effects of three internal standards.
Methods
Testosterone with two deuterium (D2), five deuterium (D5) and three carbon 13 enrichment (C13) were separately assessed. Samples were extracted using ether following the addition of 10 μL of internal standard. All aliquots were prepared in triplicate, one for each type of internal standard. After mixing, the ether was transferred to a 96-deep well block, and then evaporated to dryness. Extracts were reconstituted with 50% mobile phases and analysed using a Waters Acquity UPLC and Quattro Premier tandem mass spectrometer. This method had previously been shown to have excellent agreement with a reference method using the D2 internal standard and this was considered the target.
Results
Lower results were obtained when using D5 testosterone when compared with D2 testosterone. The C13 internal standard also gave lower results, but was closer to the D2 target than the D5 internal standard.
Conclusions
The choice of internal standard alone can have a significant affect on the results obtained by LC-MS/MS assays for testosterone using this chromatography. The effects of the combination of chromatography and internal standard choice should be investigated during method development.
Introduction
Testosterone measurement by liquid chromatography tandem mass spectrometry (LC-MS/MS) is now accepted as the preferred technique for the analysis of serum testosterone in both men and particularly women. Despite this, variation can be seen between LC-MS/MS assays on external quality assessment schemes and is most likely to be due to method differences between laboratories. 1 One area of inconsistency among routine LC-MS/MS assays is the choice of internal standard. The efficacy of an internal standard is particularly important to compensate for matrix effects during testosterone analysis as most assays involve either liquid–liquid or solid phase extraction, the efficiency of which can vary between methods. We investigated the effect of three different internal standards on the results obtained by a routine LC-MS/MS assay for both men and women.
Methods
The sample preparation was adapted from Gallagher et al. 2 Serum samples (n = 84, 41 women and 43 men), calibrators and quality controls (100 μL) were extracted using 0.5 mL of methyl tertiary butyl ether (Fisher Scientific, Loughborough, UK) following the addition of 10 μL of internal standard. The same serum samples, calibrators and quality controls were extracted in triplicate simultaneously, each replicate using a different internal standard. After mixing for four minutes, the organic layer was transferred to the wells of a 96-deep well block, and then evaporated to dryness. Extracts were reconstituted with 100 μL of 50% mobile phases (water or methanol containing 2 mmol/L ammonium acetate and 0.1% formic acid) and analysed using a Waters ACQUITY UPLC (Waters, Manchester, UK) and Quattro Premier tandem mass spectrometer (Waters). Chromatography was performed on a Synergi Hydro-RP 50 × 3 mm column (Phenomenex, Macclesfield, UK) with isocratic elution using 70% organic mobile phase. This method had previously been shown to have a minimal ion suppression profile and gave excellent agreement with a reference method using D2 testosterone (QMx, Thaxted, UK) as internal standard. 3 Therefore, the results obtained using D2 testosterone were considered to be the target concentrations of the serum samples. The precision of this assay was 11%, 7%, 6%, 5% and 4% at concentrations of 0.6, 2, 4, 17 and 29 nmol/L, respectively, to cover both the male and female ranges.
The D2 testosterone-1,2 internal standard (99% isotopic purity) was used at a concentration of 35 nmol/L in methanol; this was consistent with the published method. 2 The alternative internal standards tested were testosterone with either five deuterium testosterone-2,2,4,6,6 (D5, 98% isotopic purity) (Sigma, Poole, UK) or three carbon thirteen (C13, 99% isotopic purity) enrichment testosterone-2, 3, 4 (Isosciences, King of Prussia, PA, USA). These were prepared to the same concentration.
The primary transitions monitored in electrospray ionisation positive mode were m/z 289.2 > 96.9, 291.2 > 98.9, 292.2 > 99.9 and 294.3 > 99.9 for testosterone, D2 testosterone, C13 testosterone and D5 testosterone, respectively. The qualifying transition for testosterone was m/z 289.2 > 109.8.
Results
Lower results were obtained when using D5 testosterone compared with D2 testosterone. The Passing–Bablok regression equation for all samples was: testosterone (D5) nmol/L = 0.84 × testosterone (D2) + 0.05. The Bland–Altman agreement is shown in Figure 1. Similar regressions were obtained for men and women. For the male group, the Passing–Bablok equation was: testosterone (D5) nmol/L = 0.83 × testosterone (D2) + 0.08. In the female group, the comparison was: testosterone (D5) = 0.86 × testosterone (D2) + 0.04.
Bland–Altman bias plots using D5 testosterone (top) and C13 (bottom) testosterone compared with D2 testosterone. Analysis was of 84 serum samples: 41 women and 43 men
The C13 internal standard also gave results that were lower than the D2 internal standard target values, although they were closer to these target values than those results obtained using the D5 standard. The Passing–Bablok regression analysis was: testosterone (C13) nmol/L = 0.94 × testosterone (D2) – 0.0. The Bland–Altman agreement is shown in Figure 1. Similar regressions were obtained for men and women. For the male group, the Passing–Bablok equation was: testosterone (C13) nmol/L = 0.94 × testosterone (D2) + 0.03. In the female group, the comparison was: testosterone (C13) = 0.90 × testosterone (D2) + 0.02.
Discussion
These results show that the choice of internal standard alone can have a considerable affect on the results obtained by the LC-MS/MS assay for testosterone used routinely in our laboratory. This effect appears to be constant within the male and female concentration ranges. It is unknown how the impurities contained in these internal standards may affect the results obtained. The difference in results obtained may be due to the differing abilities of the internal standards to compensate for causes of variation such as matrix effects. If this is the case, then the effects may be different using alternative columns and chromatography and it should be possible to eliminate it completely. Despite this, it would be prudent for LC-MS/MS users to investigate how different internal standards may behave during the method development process. The D2 internal standard has been shown to give results close to the reference method using the conditions described here, but this may not give the best results using different sample clean-up procedures and chromatography columns. Investigators should be aware of these sources of variation when developing LC-MS/MS methods for testosterone.
DECLARATIONS