Carbon isotope ratios of endogenous steroids found in human serum-method development, validation, and reference population-derived thresholds

In order to detect the misuse of testosterone (T), urinary steroid concentrations and concentration ratios are quantified and monitored in a longitudinal manner to enable the identification of samples exhibiting atypical test results. These suspicious samples are then forwarded to isotope ratio mass spectrometry (IRMS)-based methods for confirmation. Especially concentration ratios like T over epitestosterone (E) or 5 alpha-androstanediol over E proved to be valuable markers. Unfortunately, depending on the UGT2B17 genotype and/or the gender of the athlete, these markers may fail to provide evidence for T administrations when focusing exclusively on urine samples. In recent years, the potential of plasma steroids has been investigated and were found to be suitable to detect T administrations especially in female volunteers. A current drawback of this approach is the missing possibility to confirm that elevated steroid concentrations are solely derived from an administration of T and cannot be attributed to confounding factors. Therefore, an IRMS method for plasma steroids was developed and validated taking into account the comparably limited sample volume. As endogenous reference compounds, unconjugated cholesterol and dehydroepiandrosterone sulfate were found suitable, while androsterone and epiandrosterone (both sulfo-conjugated) were chosen as target analytes. The developed method is based on multi-dimensional gas chromatography coupled to IRMS in order to optimize the overall assay sensitivity. The approach was validated, and a reference population encompassing n = 65 males and females was investigated to calculate population-based thresholds. As proof-of-concept, samples from volunteers receiving T replacement therapies and excretion study samples were investigated.

Automated summary

A new method for detecting testosterone (T) usage was developed and validated using plasma steroid concentrations, with a focus on female volunteers. The method may be more effective than urine samples in providing evidence of T administrations, taking into account genetic and gender factors.