Measurement of amniotic fluid steroids of midgestation via LC–MS/MS

• LC–MS/MS was validated for use of AF steroid analysis at midgestation.
• The male fetus shows higher AF levels of androstenedione and testosterone.
• LC–MS/MS data was subjected to cross-validated logistic regression analysis.
• Testosterone in AF yielded accurate predictions for male gender at birth.
• Additional analysis of further steroidal hormones did not improve this prediction.

IntroductionAnalysis of steroids by mass spectrometry (MS) has evolved into a reliable tool for the simultaneous detection of multiple steroids. As amniotic fluid (AF) and fetal serum composition of early pregnancy are closely related, the analysis of AF can yield information on the physiological status of the developing fetus. We evaluated the use of liquid-chromatography tandem mass spectrometry (LC–MS/MS) for AF steroid analysis, including the analysis of its sensitivity and accuracy for gender verification in healthy subjects.Materials and methodsAF of 78 male and 94 female healthy newborns was analyzed by LC–MS/MS at 16 weeks of gestation. The levels of androstenedione, corticosterone, cortisol, cortisone, deoxycorticosterone, 11-deoxycortisol, dehydroepiandrosterone (DHEA), dehydroepiandrosterone-sulfate (DHEA-S), 17-hydroxyprogesterone, progesterone (17-OHP) and testosterone were measured. Steroid levels were compared to RIA and GC–MS levels of midgestation from the literature. Cross-validated logistic regression was used to obtain statistical predictions of gender at birth from testosterone and the above steroids.ResultsLC–MS/MS analysis of AF steroids yielded comparable results with published GC–MS data. Gender specific differences were found for androstenedione and testosterone concentrations with higher levels in the male fetus. In contrast to published RIA data no gender specific differences were observed for 17-hydroxyprogesterone and dehydroepiandrosterone AF concentrations. Testosterone concentrations yielded highly accurate predictions for male gender at birth. Additional analysis of further steroids did neither increase the accuracy, sensitivity nor specificity of this prediction. The estimated optimal cut-off value for amniotic testosterone level was 0.074 μg/L for healthy male newborns.ConclusionsLC–MS/MS is a reliable method for the determination of steroids in amniotic fluid. The determination of testosterone in amniotic fluid by LC–MS/MS in early pregnancy of healthy subjects can be used to offer a reliable prediction of fetal gender at birth.