The analysis of sports samples for prohibited substances began in the 1960s and has developed since then using modern technologies close to the latest scientific discoveries. In this chapter the latest techniques and applications are described as well as the role of the World Anti-Doping Agency as the controlling body for the implementation of these tests. For small molecules, apart from the routine use of GC-MS, the newer techniques include the use of isotope ratio MS to detect testosterone and nandrolone administration and LC-MS/MS (liquid chromatography-tandem MS) to detect diuretics. For large molecules, several applications of LC-MS/MS are described as well as immunoprocedures for erythropoietin and human growth hormone. Finally, the latest method to detect homologous blood transfusion is briefly described.
The testing for drugs misused in sport began in the 1960s following the death of the cyclist Knud Jensen at the Rome Olympics in 1960 and other cyclists, including the British athlete Tommy Simpson, in the Tour de France in 1967. These deaths were linked with the misuse of amphetamine stimulants by these riders. The IOC (International Olympic Committee) established a Medical Commission with a role that included setting up an anti-doping programme.
At that time, advances in the detection of stimulant drugs using GC with packed stationary-phase columns and flame-ionization detection enabled the detection of analytes at concentrations in urine of the order of 1 μg/ml that were sufficient to identify the use of these substances. The development of reliable nitrogen-selective detectors, by increasing selectivity, helped to improve the ability to detect smaller concentrations of stimulants in urine. However, it soon became clear that anabolic steroids were also being misused in sport and there was a need to detect the much smaller concentrations (of the order of tens of ng/ml) of these substances in urine. The first method that was developed relied on immunoassays. Subsequently, in the 1980s, developments in GC-coupled mass spectrometers enabled more ready detection of anabolic steroids at these small concentrations and today GC-MS is the methodology of choice for detecting these substances.
In 1999, the WADA (World Anti-Doping Agency) was established by the IOC “to promote and co-ordinate the fight against doping in sport internationally”. The agency has equal representation from the Olympic Movement and public authorities. WADA produced a series of uniform anti-doping rules, the World Anti-Doping Code, which was adopted in 2003 as the basis for the fight against doping in sport. WADA documents are readily downloadable via their website (http://www.wada-ama.org).
WADA also took over the role of accrediting laboratories from the IOC. In order to be accredited by WADA, a laboratory first needs to be accredited to ISO 17025 as a testing laboratory. The WADA accreditation process is described in the International Standard for Laboratories and is a complex process that typically takes several years to complete. Once accredited, the laboratories still have to show compliance with the standard and are subjected to a series of proficiency tests four times each year as well as receiving blind proficiency samples. There are currently 33 laboratories accredited by WADA. Nineteen of these are based in Europe and analysed just over half of almost 200000 samples collected in 2006 (see Figure 1).
This chapter illustrates some of the recent successes and current problems still to be overcome in the fight against doping in sport.
WADA list of prohibited substances
WADA produces a list of prohibited substances annually . Currently this list is divided into three main categories: prohibited substances, prohibited methods and substances prohibited in particular sports. Different substances are prohibited ‘in-competition’ and ‘out-of-competition’.
The prohibited substances category is subdivided into nine sub-categories: anabolic agents, hormones and related substances, Β2 agonists, agents with anti-oestrogenic activity, diuretics and other masking agents, stimulants, narcotics, cannabinoids and glucocorticosteroids. This final category is probably more correctly named as glucocorticoids or corticosteroids.
The prohibited methods category is subdivided into: enhancement of oxygen transfer, chemical and physical manipulation and gene doping. Currently there is no recognized technique to detect gene doping.
Substances prohibited in particular sports
This final category is subdivided into alcohol and Β-blockers, both of which are prohibited in-competition only. Alcohol is prohibited in, for example, motor sports, archery and shooting where the use of alcohol could be dangerous for the competitor and/or the spectators. Β-blockers are prohibited in, for example, motor sports and aeronautics. In archery, however, Β-blockers are currently prohibited both in- and out-of-competition.
Findings of prohibited substances by WADA-accredited laboratories
Typically 1–2% of samples tested are found to contain one or more prohibited substances. The commonest laboratory findings over the years are the anabolic androgenic steroids. Of these, nandrolone and testosterone are the most frequently reported finding. Recently WADA reduced the reporting threshold for testosterone and this has resulted in an approx. 3-fold increase in reported findings although not with a corresponding increase in sanctions. This is probably, at least in part, due to the fact that the reduced reporting threshold includes part of the normal range found in the population not administering testosterone, a testosterone precursor, or a substance that affects testosterone.
The mere finding of a foreign substance or of the diagnostic metabolite of a foreign substance is sufficient evidence to bring a case against an athlete. This is because sport adheres to a strict liability policy, similar to that of drink-driving cases, i.e. it is the responsibility of the athlete to ensure that a prohibited substance does not enter his or her body. Clearly simply the finding of the presence of a substance would not be appropriate for an endogenous substance such as testosterone. There are several substances that are virtually identical with that produced endogenously that can conveniently be called ‘pseudo-endogenous’. Other examples of pseudo-endogenous substances include hydrocortisone, androstenedione, dihydrotestosterone, GH (growth hormone), hCG (human chorionic gonadotropin), and possibly nandrolone that is produced in minute quantities endogenously. In these cases, the mere finding of the substance is insufficient and instead one of the more subtle differences between the molecule administered and the endogenous molecule or a perturbation in endocrine balance must be discernable, or a concentration of the substance that so exceeds the concentration normally found in man must be determined for an offence to be prosecuted.
Detecting prohibited substances
Urine is the main sampled fluid. Blood samples were first collected at the Olympic Games in Sydney in 2000, at that time for the purpose of detecting EPO (erythropoietin) administration. The collection of blood samples does not replace the need for urine samples but enables the detection of a variety of substances that do not appear in urine. Examples include the detection of haemoglobin-based oxygen carriers, blood transfusion and GH.
The majority of substances detected are relatively small molecules of around 350 Da or less. They may be the prohibited substance itself or a diagnostic metabolite of that substance. In most cases, the prohibited substances contain one or more basic nitrogen atom and often hydroxy moieties. As such they are best analysed, following extraction from the urine sample, by chemical derivatization of the functional groups. The formation of trimethylsilyl derivatives is the derivatization method preferred by most WADA-accredited laboratories followed by GC-MS. However, in urine, many of the analytes are present as phase II metabolites conjugated to glucuronic acid. In order to extract, purify and concentrate the analytes prior to derivatization and GC-MS analysis, cleavage of the conjugate with Β-glucuronidase is required. Some laboratories prefer to use solid-phase extraction first to provide a cleaner matrix prior to the glucuronidase hydrolysis of the eluted compounds, whereas others undertake direct hydrolysis, in which case the use of a suitable deuterated drug conjugate added to the sample to indicate effective hydrolysis is desirable. The preferred source of Β-glucuronidase enzyme is from Escherichia coli since it appears to be purer, and thus free from other enzymes that might affect the analytes, compared with the older Β-glucuronidase preparations such as from Helix pomatia. Nevertheless, avoiding the hydrolysis and analysis of the intact conjugates is clearly desirable but not, as yet, routinely achievable even with the use of LC-MS/MS (liquid chromatography-tandem MS).
Furthermore, GC-MS analysis is not appropriate for molecules larger than approx. 700 Da, or for analytes that would be too unstable. LC-MS/MS is already being used for smaller molecules that do not chromatograph well through the GC column without chemical derivatization. One good example is the LC-MS/MS method developed by Goebel et al.  that enables the screening of urine samples for 35 diuretics, compounds with very varied chemical structures, with limits of detection of 100 ng/ml or less and at a rate of about five samples per h. Previously, the best method of obtaining good analytical data by GC-MS was by methylation of the compounds; this could be problematic unless excess methylating agent is excluded from the chromatograph column.
Although LC-MS/MS might seem to be appropriate, for many larger molecules such as hCG, immunoprocedures are still the method of choice. Indeed for molecules such as human GH and EPO, since they are also endogenous, the mere presence in a sample would not be evidence of administration. Recently, LC-MS/MS methods have been proposed to detect synthetic forms of insulin  and tetracosactide  in blood samples as well as for hCG  in urine. Nevertheless, it is questionable whether the mere finding of these analytes can be deemed as sufficient evidence of administration. The structure of the 24 amino acids of tetracosactide, first marketed as Synacthen, is identical with the equivalent part of the 39-amino-acid endogenous ACTH (adrenocorticotrophic hormone) molecule. Clearly the methodology used to identify tetracosactide must be able to distinguish it from ACTH. The sensitive method recently published by Thevis et al.  relies on immuno-affinity extraction from blood followed by LC-MS/MS with a limit of detection of 100 fmol/ml (0.3 ng/ml). This method relies on producing multiply charged precursor ions by electrospray ionization and producing diagnostic fragment ions by collision-induced dissociation to distinguish tetracosactide from ACTH and possible interfering components. Thevis et al. tested ten different blood samples to investigate the lack of interfering components and analysed a single patient sample 10 min post-administration of Synacthen intravenously to show sufficient sensitivity of the method. Further work will, no doubt, prove the effectiveness of this approach.
The method for the much larger glycoprotein hormone hCG (approx. 37 kDa) relies on identification of the products of a tryptic digest of the material recovered using immuno-affinity extraction of the urine sample and LC-MS/MS. The method has a claimed sensitivity of 5 m-units/ml, which is just at the WADA-specified minimum required performance limit for this prohibited substance . Nevertheless, it remains to be seen whether hCG is produced naturally in minute amounts, certainly many individuals produce small amounts of immunoreactive material and a cut-off of 10 m-units/ml had previously been proposed , following the analysis of some 1400 men, to deal with this situation. Interestingly, only recently has there been agreement [8,9] that 5 m-units/ml is the appropriate threshold to be used to detect pregnancy in females.
When chromatography and MS is used to identify a substance, WADA  requires the laboratory to establish certain criteria that include retention-time matching with reference material or a reference collection sample as well as matching of the mass spectra. For example, retention-time must agree within 1% or ±0.2 min whichever is the smaller and, in general, at least three diagnostic ions must agree within 10% or less depending on the relative abundance of the ion. These criteria are typically as stringent or more stringent than many international regulatory standards. However, when immunoprocedures are employed, the requirement is less clearly stated in the WADA International Standard for Laboratories  (paragraph 126.96.36.199.1.3) as “the immunoassay used for confirmation must use a procedure with a different antibody that should recognise a different epitope of the peptide/protein than the assay used for screening”.
Samples are normally received by the laboratory as coded duplicates, labelled as A- and B-samples, from each competitor tested. Only the A-sample is opened and its contents analysed. Should a prohibited substance be found in the A-sample, the competitor then has the right to witness the opening of their B-sample and for the analysis to take place in the presence of their witnesses including a scientific representative and/or legal representation as well as being present themselves. Despite the strictly controlled process, probably because of the strict liability principle imposed by sport, legal challenges either to the sample collection process or the analysis frequently occur. However, it is unusual for the athlete to be acquitted.
IRMS (isotope ratio MS)
Steroids in urine are profiled in sports drug testing to assist in determining whether abnormalities may be present indicating the possible “administration of testosterone or one of its precursors androstenediol, androstenedione, dehydroepiandrosterone or a testosterone metabolite, dihydrotestosterone or a masking agent, epitestosterone” . Whenever a laboratory finds a ratio of testosterone to epitestosterone greater than four in a sample, this will be reported. A longitudinal profile of the athlete must then be undertaken unless “an additional reliable analytical method (e.g. IRMS)”  has provided evidence of a prohibited substance of exogenous origin. Since the concentration in urine is rather variable because of the approx. 100-fold range of urine production, the use of a ratio of testosterone to epitestosterone helps to compensate for dilution when assessing whether testosterone or a precursor may have been administered. Nevertheless, WADA has set maximum concentrations above which a sample should be submitted to IRMS analysis. In order to be able to achieve some level of comparability between samples, WADA also requires that any concentration reported should also be reported with the concentration adjusted for the specific gravity of the urine sample, as follows:
The more definitive approach is to use carbon IRMS to assess the isotope signature of the prohibited substance or metabolite and compare this with the endogenous carbon isotope signature of the individual rather than using a population-based approach. Carbon isotope ratios are expressed in delta (δ) units per mil (%0; per thousand) against an international reference standard (δ13CVPDB) calculated from the 13C/12C molar ratio of the sample and of the international (VPDB) standard as shown:
A difference of three delta units or more of the measured steroid from that of the urinary reference steroid chosen is reported as consistent with the administration of the steroid .
Recently Hebestreit et al.  have shown that IRMS may be used to determine the carbon isotope ratio of 19-norandrosterone, one of the main metabolites of the anabolic steroid nandrolone, at concentrations perhaps as small as the WADA reporting threshold of 2 ng/ml although their method requires a 10 ml sample of urine. At this small concentration there have been legal disputes as to whether the steroid has been produced endogenously rather than from administration of nandrolone or a related prohibited anabolic androgenic steroid. IRMS may provide firm evidence that an administration has taken place. However, IRMS is unlikely to be used routinely until several of the remaining difficulties have been overcome, especially the poor sensitivity of the technique and the need for sample clean-up which is often lengthy and cumbersome, for example including the use of LC with fraction collection. Furthermore, some substances have carbon isotope signatures very similar to the normal endogenous range. Thus it is possible that a non-significant difference between sample and urinary (endogenous) reference steroid may be obtained despite the administration of a prohibited substance. Indeed, at least in theory, since it is neither a simple nor a rapid undertaking, an individual could adjust their endogenous carbon isotope signature by selecting their diet.
Detection of protein/peptide hormone and homologous blood administration
Tests for the detection of the administration of the glycoprotein hormone EPO were first adopted internationally at the 2000 Olympic Games in Sydney. Two tests were adopted, one based on the detection of the EPO isoforms [14,15] distinguishing the normal endogenous profile from that of the recombinant forms of EPO (rHuEPO) that are commercially available. The other approach, which requires a blood sample, evaluates blood parameters as indirect markers to detect altered erythropoiesis caused by administration of rHuEPO . The test is based on multivariate statistical models; one is the ‘ON’ model that uses the reticulocyte haematocrit, serum EPO, soluble transferrin receptor, haematocrit and percentage of macrocytes; the other is the ‘OFF’ model using reticulocyte haematocrit, serum EPO and haematocrit. The author suggests that both tests be applied to provide the best detection both during EPO administration and for the longest period thereafter, perhaps as long as 4 weeks. Although this test has been refined over the last few years, it has not as yet been widely adopted perhaps because the isoform test only requires a urine sample. The isoform test developed by Lasne and colleagues [17,18] is based on an immuno-electrophoretic approach using double-blotting to reduce non-specific binding of secondary antibodies. However, because EPO has such a short half-life, the test can detect rHuEPO use probably for a maximum of about 3 days post-administration. This isoform test was successfully used at the Winter Olympic Games in Salt Lake City in 2002 to disqualify three gold medallists who had used darbepoietin, a more heavily glycosylated synthetic analogue of rHuEPO  that has a much longer half-life than rHuEPO. Unfortunately, the isoform test suffers from requiring many manipulations and takes approx. 1.5-3 days to complete. Also, Beullens et al.  recently claimed that the test can occasionally lead to false-positive identification of rHuEPO in urine post-exercise because of a lack of specificity of the monoclonal anti-EPO antibodies used. This has been strongly contested by Catlin et al.  and by Lasne , and WADA have a strict policy in place to ensure that EPO data are carefully and independently checked before a result is released.
Wu et al.  have developed a test to detect recombinant human GH administration (rHuGH) based on the detection of the 22 kDa only isoform present in rHuGH against the mix of isoforms, especially 20 and 22 kDa, of endogenous pituitary-derived GH. However, since the half-life of GH is very short, the detection period is likely to be similarly short and, to date, no athlete has been prosecuted for GH administration using this method. Sönksen et al.  have been researching a method that makes use of biomarkers, especially IGF-1 (insulin-like growth factor 1) and PIIIP (procollagen III peptide), that should have a greater detection interval than the isoform method.
Homologous blood transfusion
Nelson et al.  have recently developed a test to detect the transfusion of homologous blood. Although when blood transfusions take place care is taken to ensure that the main blood groups ABO and Rh(D) are matched, the transfused blood is not generally matched against minor blood group antigens. Thus the test is based on appropriate panels of antibodies to these minor blood-group antigens. The test claims to be able to detect less than 5% of transfused blood, that is as little as half of one unit of blood, with a good detection period since transfused blood is known to remain in the circulation for at least 90 days against the lifetime of a red blood cell of about 120 days.
This chapter has attempted to illustrate the recent scientific successes that have facilitated the control of doping in sport. A general review on the analytical aspects of the subject has been published by Trout and Kazlauskas . Improvements to the sensitivity of detection methods continue and research into ‘designer’ drugs will limit the possibility of evasion of detection. Work on the detection of the potent protein hormones has already proven to be successful and further research, especially on GH, is likely to soon provide improved testing methods. Concerns over the detection of possible gene doping have not been addressed but further research in this area is essential. The move from using reference populations to individual reference values, in the form of a so-called athlete passport, will be a crucial step in providing a more discriminating method to detect cheats and to enable clean athletes to compete fairly. Unfortunately, sports drug testing is relatively slow compared with other methods used to ensure fair sport and it is likely to remain so until point-of-collection analytical devices have been developed to a point where they may have sufficient certainty to disqualify a sports competitor.
• Methods exist to detect the administration of: small foreign molecules such as synthetic anabolic steroids, stimulants, diuretics and masking agents; testosterone and testosterone precursors using the testosterone/epitestosterone ratio as well as IRMS; protein hormones including hCG and EPO; and homologous blood transfusions.
• Research is underway to improve detection of rHuGH.
• Development of individual reference ranges is important.
• Development of a methodology to detect gene doping is needed.
• Development of point-of-collection analytical devices are needed.
- © The Authors Journal compilation © 2008 Biochemical Society