The most abused substance in sports testing is testosterone, an anabolic steroid that has been on the International Olympic Committee's list of banned substances for many years. It is regularly found in urines tested in official labs accredited by the World Anti-Doping Agency (WADA), despite the best efforts of sportsmen and sportswomen to disguise their actions.

In practice, testosterone itself is not the drug taken, because it is metabolised rapidly to inactive compounds. The preferred alternatives are testosterone esters, which metabolise slowly to testosterone, providing a more gradual release of the steroid into the system. For drug testing labs looking for testosterone esters during the routine testing of urine samples, the focus is on the metabolite testosterone, because the esters are not detectable in urine.

In order to distinguish the drug from natural testosterone, the ratio of testosterone to epitestosterone, a second natural steroid, is measured, although the proximity of the ratios from natural and abused testosterone has caused problems in the past. It would be far preferable to measure the intact testosterone esters.

The esters have been found in blood plasma but a less invasive matrix for collection would be hair, and this has been investigated by scientists in Belgium. Oscar Pozo, Koen Deventer, Peter Van Eenoo and Frans Delbeke from the DoCoLab at the University of Ghent, a WADA-accredited laboratory, and Robert Rubens from the State University Ghent Hospital devised a method for testosterone undecanoate, a common ester, using liquid chromatography/mass spectrometry.

They tested hair taken from three patients who had been prescribed the drug and used negative hair samples from healthy volunteers for controls and validation studies. Separation was carried out on a C₈ column with a mobile phase gradient of methanol in aqueous formic acid-ammonium acetate. Under these conditions, the retention times of testosterone undecanoate and the internal standard testosterone decanoate were less than 7 minutes.

With solutions of standard steroids, electrospray ionisation (ESI) was more sensitive than atmospheric pressure ionisation (APCI) but this was reversed for real samples. The change was attributed to differences in background noise and ion suppression which made APCI the preferred technique.

Despite this choice, there was still significant background noise, which varied between samples. Most of the noise arose from a peak in the extracted ion chromatogram on the shoulder of the m/z 457 peak for testosterone undecanoate. Closer examination revealed a peak corresponding to a sodium adduct at m/z 457 which was attributed to poly(ethylene glycol), a common ingredient of shampoos and hair products.

Thorough washing of the hair samples with methanol removed much of the contaminant and reduced noise levels to close to those of the standard. Subsequent digestion with tris(2-carboxyethyl)phosphine and extraction with pentane gave acceptable recoveries of 83% from spiked samples.

The APCI mass spectra of testosterone undecanoate and the internal standard testosterone decanoate gave strong protonated molecular ions and good fragmentation. Four fragment ions from the drug and two from the standard were earmarked for selected reaction monitoring studies.

Detection limits of the optimised method were 0.2 ng/g, which is far better than those of published GC/MS methods which range from 1-20 ng/g. Accuracies were 92-102% at 0.5-100 ng/g and the relative standard deviations were less than 7%.

The three patients all took daily capsules containing 40 mg of the drug but the levels of testosterone undecanoate in their hair were very low at 0.4, 1.6 and 8.4 ng/g. The data illustrate the necessity for a sensitive testing method.

Hair testing of athletes has the bonus that it will be far more difficult to falsify samples, as occurs with urine testing. It would be very difficult for an athlete to refuse permission for an official to remove a few hairs from their head on the basis of privacy. On the other hand, head shaving could be an avoidance technique, so the method would need to be tested on other body hair.

The next step, say the researchers, is to carry out a broader study with more samples and to check the distribution of the drug along individual hairs, which would help to confirm long-term drug use.

Related links:

• Biomedical Chromatography 2009, 23, 873-880: "Quantification of testosterone undecanoate in human hair by liquid chromatography-tandem mass spectrometry"

Article by Steve Down

The views represented in this article are solely those of the author and do not necessarily represent those of John Wiley and Sons, Ltd.