When he worked for the California-based biotechnology company Amgen in the 1990s, Steve Elliott was instrumental in the development of darbepoetin alfa (Aranesp), a drug to treat anemia in people with chronic renal failure. Approved by the US Food and Drug Administration in 2001, darbepoetin is a second-generation version of recombinant human erythropoietin (EPO) purified from Chinese hamster ovary cells that express the modified gene.
EPO is a cytokine produced by the kidneys that sends a signal to the bone marrow to increase production of red blood cells (RBCs). For anemic patients, the drug can be lifesaving. But for people with normal red blood cell counts, it offers something else: a boost in physical performance by increasing the amount of oxygen the blood can carry to muscles. As Elliott his colleagues developed the drug, “it was apparent that Aranesp would be used for doping,” he...
Sure enough, after darbepoetin was approved by the US Food and Drug Administration (FDA), it became a tool for performance enhancement among some athletes. Possibly assuming that a test to detect it would not yet be available at the 2002 Winter Olympics in Salt Lake City, Utah, some athletes dosed themselves with darbepoetin leading up to and during the Games. But there was a test. Elliott and his colleagues had developed the detection method in parallel with the drug’s coming to market, but they did not publicly announce the test’s existence prior to the start of the Games.
Compared to recombinant human EPO already on the market for medical use, the protein that Elliott and his colleagues had engineered had a longer half-life, allowing for less-frequent dosing of anemia patients. This feature also gave anti-doping enforcers a leg up, because the molecule lingered for longer in the body, giving them more time to detect it. Moreover, the unique combination of sugar molecules on the surface of darbepoetin distinguished it from endogenous EPO and from other recombinant EPOs.
In 2002, an anti-doping lab at the University of California, Los Angeles, deployed the urine-based test, which could detect darbepoetin’s distinct glycosylation profile. At that year’s Winter Olympic Games, a local lab used the test to analyze athletes’ samples, and late in the Games, three skiers—Johann Mühlegg of Spain and Olga Danilova and Larissa Lazutina of Russia—tested positive for darbepoetin and were disqualified.
A common misconception is that when an athlete’s sample is collected that it’s ana-lyzed with a simple dipstick and that it’s either positive or negative.—Matt Fedoruk, US Anti-Doping Agency
But the number of athletes charged with darbepoetin doping dropped off quickly after that, as some athletes switched to blood transfusions or started repeatedly taking much smaller amounts of darbepoetin, called microdoses, to keep levels below detectable limits—a strategy they already employed for other forms of recombinant EPO. Another approach to hide their drug use was to infuse recombinant EPOs directly into circulation, rather than injecting it below the skin, so that the molecules are cleared more quickly from the body. Athletes also started supplementing their recombinant EPO molecule of choice with iron, which is a key component of hemoglobin. The additional iron increases the effectiveness of the EPO, meaning that an athlete can take lower doses. The iron also masks EPO-related biomarkers that are used as indirect tests of EPO use.
Today, “we see that no one is taking just erythropoietin,” says Michael Pearlmutter, executive director of the nonprofit research collaborative Partnership For Clean Competition (PCC), for which Elliott now consults. “They are combining it with iron or other substances, which makes the doping more difficult to detect.” The problem for athletes—an opportunity for anti-doping labs—is that taking iron alters blood-borne biomarkers of EPO use that officials are now testing for alongside the various forms of the recombinant hormone itself.
And so continues the cat-and-mouse game between athletes and anti-doping officials. As athletes experiment with drugs that can help give them an edge over their competition, they tweak their strategies to try to evade detection. Meanwhile, anti-doping researchers are continuously developing assays and other methods that are more sensitive to low levels of doping agents that accumulate in urine or blood, or that extend the detection window to secure a positive test from a cheating athlete.
“Athletes are continuing to become more sophisticated in their doping efforts,” says Pearlmutter. “It’s the reason why the PCC exists. If doping were static then anti-doping could be static.”
Establishing a baseline
For athletes who want to cheat, the most attractive substances are those that are already found in the body. Besides EPO, these include growth hormone, steroids such as testosterone, and blood transfusions. Because exogenous and endogenous biomolecules can be challenging to distinguish and because there is a wide range of “normal” levels for these compounds and their associated metabolites, these agents present a particular challenge to anti-doping watchdogs.
“For most of the tests, there is quite a lot of variability in the biomarkers from one individual to another,” says endocrinologist Richard Holt of the University of Southampton in the UK.
To make the testing results for steroids and other doping agents less ambiguous and to allow for more effective and sensitive monitoring, the World Anti-Doping Agency (WADA) introduced the Athlete Biological Passport (ABP) to track biomarkers in individual athletes over the long term. In 2009, WADA formally launched the ABP’s blood module, which tracks markers of blood doping in circulation. Levels of hemoglobin, red blood cells, and other blood-borne markers of EPO and whole-blood doping are monitored. In urine samples, peptide hormones, including growth hormone, are also tracked.
Five years later, WADA added the steroid module to the ABP: urine tests to specifically check for testosterone and other steroid doping by measuring not only the steroids themselves, but also immediate precursors or derivatives, and secondary molecules such as metabolites that fluctuate with steroid levels. An athlete’s passport is flagged for further evaluation, and possibly more testing, when levels of these biomarkers are outside of the athlete’s 99.5th percentile reference range, calculated with a mathematical model that is based on a population of healthy individuals and an individual’s own results. As an athlete accumulates more test results, her individual reference range narrows.
“A common misconception is that when an athlete’s sample is collected that it’s analyzed with a simple dipstick and that it’s either positive or negative,” says Matt Fedoruk, scientific director at the US Anti-Doping Agency (USADA). “Rather, we are collecting blood and urine, and there is a high level of sophistication that is applied to detect hundreds of prohibited substances. The analyses are quite challenging and complex.”
Anti-doping researcher Jenny Mullen and colleagues at the Karolinska Institute in Sweden have demonstrated that, by establishing a baseline, the ABP could enable detection of a single 100-milligram dose of testosterone gel to the skin—what would be considered a microdose—as long as seven days after application in healthy male volunteers.1 In another study, Mullen’s team found that injection of a 125-milligram microdose of testosterone enanthate, a drug intended to boost levels of testosterone among males with low levels of the steroid, could be detected using the same methods for as long as eight weeks.2
“WADA has rightly set the bar very high for [doping] detection, but that means we don’t always have the sensitivity we would like for our tests,” says Holt. “The ABP is a mechanism by which athletes who travel around the world to compete can be tested by different labs, and those results are comparable among those labs and centrally stored as part of the athlete’s history.”
The Doping-Detection Passport
The World Anti-Doping Agency recently developed and instituted the Athlete Biological Passport (ABP), a tools that helps anti-doping organizations track levels of various molecules in competitive athletes. In 2009, WADA published formal guidelines on how to conduct standardized testing for evidence of blood doping, including how to track athletes’ longitudinal data as part of the ABP’s blood module, which includes 14 bloodborne biomarkers to indicate misuse of erythropoietin (EPO), blood transfusions, or other forms of doping using blood samples.
In 2014, WADA added the steroidal module to track the levels of testosterone and other steroids. These sets of tests can detect exogenously administered steroids, their various metabolites, precursors, and related molecules to better nail down doping as well as other anabolic agents. Tracking these markers over time is also a way to identify samples that may have been tampered with or exchanged with the urine sample of another individual, as happened with the Russian Olympic team during the 2014 Winter Olympics in Sochi.
In addition to tracking biomarker levels in athletes over time, anti-doping officials are working to understand what the relative levels of various testosterone precursors and related markers can tell them about an athlete’s normal physiological levels and doping status. Such comparisons are critical to the identification of exogenous testosterone use: numerous precursors of the steroid hormone, as well as epitestosterone, an isomer of testosterone, are suppressed in the presence of exogenous testosterone. Precursor levels relative to the levels of testosterone in the body over time can indicate if something unnatural is going on. In 2007, researchers devised a statistical technique to detect individuals’ abnormal steroid values,3 and WADA added the steroid module to the ABP in 2014.
But there can still be some ambiguity as to what these test results mean. If the variability of steroid biomarkers or their ratios are suspicious to the anti-doping scientist doing the assessment at a WADA-certified lab, or if steroid biomarker values are too high based on set rules set by WADA, the testing authority can order a pricier but definitive test that combines mass spectrometry and gas chromatography to measure the ratio of two carbon isotopes, carbon-12 and carbon-13, in urine samples. These carbon isotopes in endogenous testosterone come from the individual’s diet and differ from the carbons found in lab-made hormone, which has a lower carbon-13:carbon-12 ratio. If an athlete’s sample exceeds WADA-determined thresholds, they are considered to have tested positive for testosterone doping.
Anti-doping watchdogs measure other biomarkers that can influence an athlete’s urine levels of steroid hormones to reach stronger conclusions on the likelihood of doping. For example, an individual’s alcohol consumption or the presence of bacteria in the bladder or in the stored sample can cause an increase in testosterone levels, so labs look for bacteria and alcohol-specific metabolites in urine. And most recently, scientists involved in a PCC working group rolled out a liquid chromatography– and mass spectrometry–based test that achieves greater sensitivity in detection of testosterone doping with repeated testing of the steroid’s levels in blood serum, rather than urine. This test, administered by WADA-accredited laboratories just before the 2016 Olympic Games in Rio de Janeiro, caught two female Ukrainian track and field athletes who, the Court of Arbitration for Sport confirmed in April, had illegally used testosterone.
Still, researchers know they need to constantly improve tests and testing strategies. “We need to be aware that athletes will continue to adapt their doping behavior; therefore, ongoing efforts to develop and refine new detection strategies are of utmost importance,” says Fedoruk.
To have a conclusive test, we need to make sure the markers we are measuring are not affected by training and changes in diet, injury, or pathology.—Richard Holt, University of Southampton
Adaptability is also the strategy for officials deciding which Olympics-bound athletes to test for doping agents and how often. “All testing is intelligence-based and strategically-driven to maximize the deterrence and detection of performance-enhancing drugs,” says Fedoruk. The USADA relies on recent doping trends, the history of doping in a specific sport, the ABP, and tips received on possible doping situations to determine the tests athletes undergo.
Another naturally occurring molecule abused by athletes is human growth hormone, which naturally increases in the body as a result of exercise, can make muscles grow, and allows faster recovery after exertion. If taken during training, recombinant growth hormone can both decrease fat and increase lean body mass. By the end of the 1980s, growth hormone was among the substances prohibited by the International Olympic Committee, yet scientists only launched the test to detect the compound in 2004.
The majority of the endogenous protein that circulates in the body is 22 kilodaltons in size, but an alternative splicing pattern of the messenger RNA can result in a 20-kilodalton form. The lab-made product consists of only the 22-kilodalton molecule, and taking exogenous growth hormone blocks the natural production of growth hormone by the pituitary gland, explains Holt. “[This] changes the ratios of the various isoforms, which can be picked up by the current assays.”
But interpreting the ratio has its challenges. Although each individual has a stable ratio of the isoforms, and deviations signal that doping is likely, those deviations are short-lived because endogenous growth hormone production rapidly resumes in the pituitary gland after doping, providing only a short window of 12 to 24 hours for authorities to detect misuse. Moreover, athletes conceivably could continue to use growth hormone sourced from human cadavers, which contains all of the endogenous isoforms and cannot be picked up using the isoform test. “Judging by the relatively low number of growth hormone positives globally, further research into increasing the sensitivity of current growth hormone detection methods is warranted,” says Fedoruk.
To overcome detection challenges, Holt led the development of a second test to indicate the hormone’s misuse. Called the GH-2000 score, it relies on immunoassays and mass spectrometry methods to measure levels of two proteins—insulin-like growth factor-1 (IGF-1) and N-terminal propeptide of type III collagen (P-III-NP)—that increase following growth hormone dosing. The test is approved by WADA and was launched in the run-up to the 2012 London Olympic and Paralympic Games.
But this biomarker test hit a manufacturing setback when one of the makers of one of the assays withdrew its kit from the market. So Holt and others are now working with alternative assays to measure both IGF-1 and P-III-NP. One of the newer assays involves mass spectrometry to measure IGF-1. Approved by WADA in 2014,4 it is now routinely used in combination with P-III-NP immunoassays. But researchers are still working to develop a reliable P-III-NP mass spectrometry method.
And challenges face the existing tests. Holt and his colleagues have accumulated evidence that “neither the isoform nor the biomarker test is performing as well as we had hoped,” he says. The biomarker test has a longer window of detection than the isoform test—about one to two weeks—but levels of the biomarkers can differ widely between individuals, making it difficult to draw concrete conclusions about an athlete’s potential doping behavior. “To have a conclusive test, we need to make sure the markers we are measuring are not affected by training and changes in diet, injury, or pathology,” says Holt.
Growth hormone detection is not currently part of the ABP, but WADA is considering adding either the isoform or biomarker test despite the flaws of both. One issue is that the ABP’s steroid module measures steroids in urine, but growth hormone requires a blood draw: the molecule is larger, and therefore not readily passed through the kidneys into the urine. Blood tests are generally done less frequently, as they are less athlete-friendly than easier-to-administer urine tests.
“These tests are expensive, on the order of $50 a test,” says Holt, “and so it’s not a trivial decision to keep adding additional tests without demonstrating their effectiveness in detecting growth hormone misuse.”
Boosting with blood
Another performance-boosting tactic that can be challenging to detect is a blood transfusion. By increasing the volume of RBCs, athletes can carry more hemoglobin, allowing the blood to transport more oxygen to the muscles, similar to the effect of EPO. The transfusions can come from either a donor or the athletes themselves, with the former being easier to detect. In 2004, testing labs implemented a method to detect blood transfused from a donor by looking for changes in RBC populations. The test can detect antigen differences on the cells, and it caught many cheaters, including Tyler Hamilton, an elite cyclist who coauthored a tell-all of his own doping experiences.
To indirectly detect autologous blood transfusions, WADA has developed a test for plasticizers, such as di-2-ethylhexyl phthalate (DEHP), that can leach from storage bags into the blood, but plasticizer detection is not a widely used method. A major issue is that plasticizer detection is not reliable because these compounds are also found in other common products including upholstery, food packaging, and children’s toys, and can leach into the environment and ultimately into athletes’ bodies through their food and water.
As another option, Jen-Tsan Chi’s lab at the Duke University School of Medicine, with funding from WADA and the PCC, looked for changes in RNA expression as RBCs aged for up to 42 days—the longest time US blood banks allow blood to be refrigerated for transfusions. The results, published last October, revealed a distinct increase in two microRNAs and a drop in two others during these six weeks of storage.5 “The abundant and diverse species of RNA in the mature RBC offer a novel way to characterize the changes in RBC property during storage,” says Chi, who adds that the lab is now working on a single-cell approach to identify RBC microdosing.
Getting ahead of users
As researchers continue to develop more-sensitive assays, they can better equip anti-doping officials to keep athletes on a level playing field. Yet results of highly sensitive tests can also underscore gaps in knowledge on how performance-enhancing substances are metabolized and cleared from the body.
In December 2018, for example, UFC fighter Jon Jones tested positive for trace amounts of a long-term metabolite of a common doping steroid called oral turinabol (the “little blue pill” used by the East German state-sponsored doping scheme in the 1970s and ’80s), raising the issue of whether a detected metabolite is a sign of recent doping or a remnant of long past use. Jones had tested positive for turinabol back in July 2017 and had been suspended by the California State Athletic Commission for 15 months. After the December 2018 test result, the USADA concluded that because only the long-term metabolite and not the steroid itself was detected, the most likely scenario is that Jones had not ingested oral turinabol recently. Ultimately, he was allowed to continue to fight.
The tests are getting better, says Fedoruk, but then “the challenge becomes with interpretation of the findings so the results management process is fair and just for the athlete.”
Researchers continue to delve into the biology of how the body processes exogenous versions of naturally occurring hormones and other molecules in order to develop more-sensitive and more-accurate tests, Pearlmutter, Elliott, and Holt all acknowledge. And the PCC is specifically working on dried plasma and saliva-based tests, which could circumvent costly transport of temperature-controlled blood samples and the need for a phlebotomist to be present at testing sites. But details of this development are closely guarded secrets. “You obviously don’t want to tell athletes exactly what testing is coming next,” says Elliott. “Then you get behind.”
- J. Mullen et al., “Sensitivity of doping biomarkers after administration of a single dose testosterone gel,” Drug Test Anal, 10:839–48, 2018.
- J. Mullen, “Urinary steroid profiles in doping testing,” Thesis, Division of Clinical Pharmacology, Karolinska Institutet, Stockholm, Sweden, 2018.
- P.-E. Sottas et al., “Bayesian detection of abnormal values in longitudinal biomarkers with an application to T/E ratio,” Biostatistics, 8:285–96, 2006.
- J.K. Powrie et al., “Detection of growth hormone abuse in sport,” Growth Horm IGF Res, 17:220–26, 2007.
- W.H. Yang et al., “Angiogenin-mediated tRNA cleavage as a novel feature of stored red blood cells,” Brit J Haematol, 185:760–64, 2019.