The Hamilton decision

Special news feature, April 19, 2005

Is he guilty and is the science perfect?

The American Arbitration Association/North American Court of Arbitration for Sport's decision regarding the Tyler Hamilton blood doping case has set some interesting precedents and raised some important questions in regard to sanctioning athletes for drug offences, as Cyclingnews' Chief Online Editor Jeff Jones reports.

On Monday, April 18, the two out of three AAA/CAS arbitrators ruled that Tyler Hamilton should be given a two year ban from cycling after testing positive for a homologous blood transfusion (injecting someone else's blood) at the Vuelta a España on September 11, 2004. Hamilton also returned an A sample positive on August 25, 2004, after winning the Olympic Time Trial, but was not sanctioned for that as his B sample was effectively destroyed by the Athens lab that did the testing, and no result could be determined from the sample.

Hamilton's Vuelta samples were analysed by the lab in Lausanne, which determined that both A and B samples showed signs of a mixed red blood cell (RBC) population. This same lab analysed samples taken from Hamilton's former Phonak teammate Santiago Perez on October 5, a week after the Vuelta had finished. They too showed signs of a mixed red blood cell population, and Perez has been given a two year ban for blood doping by the Spanish Cycling Federation.

Hamilton's defence and USADA's attack

According to the AAA/CAS report, the Respondent, Tyler Hamilton's main points of defence in his case were:

• The test method is insufficiently validated for use in an anti-doping context;
• A quantitative as opposed to a visual standard should be used to determine a 'positive' result;
• A 'positive' result means the presence of a mixed red cell population only and does not prove homologous blood transfusion;
• There are other possible explanations for the presence of a mixed red cell population, such as disease, bone marrow transplantation, intrauterine twin-twin transfusion, and chimerism.

The claimant, the US Anti-Doping Association (USADA) took the stance that the arbitration panel had to consider a) Whether Hamilton had mixed populations for two or more blood cell markers in his Vuelta sample? and b) Whether Hamilton had given a reasonable explanation of the presence of the mixed RBC population, other than that of a homologous transfusion.

The test

Hamilton's case was the first time that the homologous blood test (HBTT) developed by the Australian Ashenden-Nelson group had been used to declare an athlete in any sport positive. The test was introduced at the Athens Olympics last year and is based on flow cytometry, which can detect mixed red blood cell populations by binding surface markers to a fluorescent tag. A histogram of the types of red blood cells present in a sample can be obtained using a flow cytometer, and the number of "peaks" seen on the histogram corresponds to the number of types of red blood cell populations present.

According to criteria set down by the World Anti-Doping Agency (WADA), there must be "a distinct peak for two different markers on the histograms before it is concluded that a blood sample contains a mixed population of RBCs." In normal medicine, only a single peak with a shoulder or tail is necessary to indicate that there are mixed populations of RBCs. "Therefore, the WADA criterion gives the benefit of the doubt to the athlete and is a conservative approach to the assessment of a doping infraction."

In the AAA/CAS report, it is stated that, "There is no division of opinion or conflicting scientific evidence that indicates any dispute between the various experts and the published scientific literature that the flow cytometry methodology can identify a mixed blood cell population. The reliability of the flow cytometer to identify two blood populations is not challenged in these proceedings."

The test works because although it is quite easy to match blood for a transfusion for the major surface markers (A, B, O and Rd(D)), it is highly improbable to match it for all the minor surface markers. The test relies on picking up the differences, if any, in the minor surface marker profiles. If there are differences, then the likely, but not the sole, conclusion is that a homologous blood transfusion has taken place.

In his defence, Hamilton argued that there should be a "lower limit of detection" for a positive result of 5% of the total RBC in the sample. USADA argued that the test is merely an identification test. According to the majority of the scientific literature, the flow cytometer can pick up mixed RBC populations to a level of 0.07%. According to the Nelson-Ashenden group that developed the test, "a mixed population of 1.5% - 2.9% is clearly distinguishable from a person's endogenous RBCs." Furthermore, the Nelson-Ashenden group observed a blood sample with a mixed population of 0.4%, which was made 111 days after the sample was drawn.

"Therefore, at any percentage level above 'noise' the test is identifying two populations or a homogenous population...The noise level of the flow cytometer used in conducting the HBTT has been calculated at 0.1%, a value that is well below the 5% threshold argued by the Respondent." Hamilton's histogram showed a threshold percentage of at least 1.3%, which was deemed sufficient to declare him positive.

No false positives

One of the most surprising statements in the AAA/CAS report, and one that was argued against by the dissenting arbitrator Chris Campbell, was that "There can be no risk of a false positive" in this test. "In fact, in the 48 subjects reported in the literature, there was 100% accuracy. There is no risk of a false positive and no need to do so called validation studies."

Do we have an example of the first unfalsifiable scientific method? "There was no need to do the validation studies" is a different statement to "validation studies were carried out and there were no false positives reported." It implies the test is perfect and cannot be questioned. Has science really come this far? Can we therefore ask WADA to approve Cold Fusion?

The test developers argued, "The markers on the RBCs of a person are the results of predetermined human genetics. The test identifies two different human blood populations based on this genetic certainty...If the HBTT identifies two different RBC populations, it stands to reason that one of the RBC populations must have come from another person who has his or her own unique, genetically pre-determined set of RBC surface markers."

This was enough to convince the majority (2 out of 3) of the arbitration panel that the test was valid.

Of vanishing twins and chimeras

Working under the assumption that Hamilton did in fact have a mixed RBC population in his body on September 11, 2004, the next part of his defence focused on other causes of said population. Both disease and bone marrow transplant were "ruled out as having a zero probability", leaving Hamilton to argue that he had a 'vanishing twin', that transferred some of its RBCs to Hamilton while he was in the womb, but disappeared in the first trimester of pregnancy; or that he is a human chimera, with a natural mixed RBC population.

The evidence states that some of Hamilton's flow cytometry histograms only show the presence of one type of RBC: one on February 5, 2005, and one taken by Dr David Housman, a genetics expert and MIT professor, which was not produced at the hearing but was claimed to show only one peak.

According to testimony given by Australian scientist Dr. Ross Brown, "It is highly unlikely that 34 years later [after being born], Tyler Hamilton would have mixed RBC populations indicated on some histograms including the one in question, and have other histograms with no mixed population indicated. Therefore, the vanishing twin phenomena can be ruled out as being extremely remote and not to have been the likely cause of the mixed RBC population."

Hamilton's other argument was that he could be a human chimera, with two different, genetically distinct RBC populations. Hamilton cited a paper by Van Dijk ("Blood Group Chimerism in Human Multiple Births", 61 American Journal of Medical Genetics 264 (1996)) that suggested chimerism can occur in up to 30% of the population. USADA and its experts argued that chimerism is too remote to be taken account of as the cause.

Hamilton's defence relied heavily on the Van Dijk paper, which "seems to be an isolated paper in the medical literature." Furthermore, Dr. Ross Brown said that the figures in the study were "just basically all so wrong, and there's so many mistakes in the paper that we basically discounted it." Dr Brown added that he had never observed a chimera in testing over 20,000 blood samples, and the Red Cross had only observed one chimera case in test millions of blood samples tested over 20 years.

Other papers suggest that there are no more than 100 cases of human chimerism in existence, making the probability that Hamilton is a chimera very small. Then there is the question of Hamilton's differing histograms, some of which show two peaks, and some of which only show one. Hamilton again relied on the Van Dijk paper to argue that a human chimera could show a mixed RBC population on one day and a single population on another. "The Van Dijk research is the only scientific study of its kind relied on by the Respondent to suggest that such fluctuations are possible in human chimeras. This assertion loses considerable strength when it is accepted that human chimerism is an extremely rare occurrence."

Similarly, a theory that Hamilton could be a microchimera, with some of his mother's RBCs present in his body, was dismissed on the basis that it could only be detected at a level that was too low to be responsible for the Vuelta results.

The verdict

The majority of the AAA/CAS panel came to the conclusion that "the mixed RBC population arising from the Vuelta sample analysis has a very high probability of having [been] caused by a blood transfusion, and an extremely low to the point of negligible probability of having been cause by Tyler Hamilton being a human chimera.

"The Panel therefore finds that a doping violation has been committed by Tyler Hamilton. the minimum suspension for a first offender in accordance with UCI Anti-Doping rule 261 is two years. Tyler Hamilton is therefore suspended from competition for a period of two years commencing, April 18, 2005. All of his competitive results from September 11, 2004, including the Vuelta competition are cancelled."

The fall out

It is noteworthy that two out of the three arbitrators agreed on the decision to declare Hamilton guilty. The third, Christopher L. Campbell, dissented on several grounds. The first, alluded to above, is that the "Lausanne Laboratory's failure to provide the measure of uncertainty means its testing method failed to meet the prevailing standards of the scientific community."

Campbell strongly asserted that "The Testing Method should have accurately calculated the rate of false positives when the presence of a second RBC population was not caused by a homologous blood transfusion." Neither the Nelson Study nor the Lausanne laboratory did this. "False positive results do not appear to be a problem," said the Nelson Study, which subsequently contradicted itself and admitted that a second RBC population could be present without a homologous blood transfusion, i.e. in the case of disease, a bone marrow transplant, a vanishing twin, or a chimera. Although none of these were deemed likely in Hamilton's case, it does not clear the test of all doubt.

The Nelson Study did not mention that false positives could also occur if a lab had not carried out the testing method properly. "This omission is odd because it was a Nelson validation study that demonstrated that in its quantitative testing method, laboratories falsely identified second RBC populations where none were present up to a level of 1.1%. The Nelson Study called these false positives background events and stated they were caused by incorrect gating and other problems. Mr Hamilton's sample in this case was within the range of a background event."

Campbell also criticised the test for being subjective. "If a test was gated in the wrong region, it would also impact a peak. This also refutes any argument that if a second peak is visible, the only explanation is a second RBC population."

Campbell raised the issue that the test would be used on women, who could have fetal cells circulating in their blood system in case of a failed conception. "Pregnancy in and of itself may establish a long-term, low grade chimeric state in the human female."

"Because the WADA Transfusion Positivity Criteria and the Testing Methods fail to provide an accurately calculated rate of false positives, it fails to satisfy the prevailing standards of the scientific community. In this situation, USADA should not be able to sustain its initial burden of proof and the case against Mr. Hamilton should be dismissed," Campbell stated.

He also went into detail about the subjectiveness of the identification method. When Hamilton's A sample in Athens was first analysed by the laboratory in Greece, it was deemed negative. "Nevertheless, on September 16, 2004, almost a month later, the IOC formed an external expert group that ruled Mr. Hamilton's Athens sample was positive. This group of experts apparently consisted of individuals paid for developing the Testing Method and individuals who knew the sample belonged to Mr. Hamilton. Two of the cardinal rules of drug testing are that the individual doing the analysis (1) should not have a vested interest in the outcome and (2) should not know the identity of the individual providing the sample."

The visual identification method, "has not been peer reviewed or properly validated," wrote Campbell. "The panel's acceptance of the deficiencies in the WADA Criteria and Testing Method establishes a dreadful precedent."

Finally, Campbell referred to Haven Hamilton's response when asked whether the statements made in the press by high ranking WADA and IOC officials gave her concerns regarding whether her husband would receive a fair hearing. "Absolutely. Absolutely. And I think that especially when Jacques Rogge, the president of the IOC, came out two days before Christmas and said Tyler Hamilton's entire career should be called into question because of this test result. That's an unbelievably horrible thing to say about someone...I just think it's inappropriate, and I don't think any athletes should be subjected to any of that, because it can't help but influence people."

Campbell concluded by saying, "Athletes should not have to worry that high ranking officials are sending clear messages to the arbitrators to find the athlete guilty regardless of the facts of the case. The IOC and WADA should consider making rules prohibiting such conduct to comply with a very important fundamental principle of the Olympic movement, fairness."

Hamilton's next step: Appeal to the CAS

In light of the above, it appears that Tyler Hamilton does have a chance to successfully appeal the AAA/CAS decision to the International Court of Arbitration for Sport in Lausanne, Switzerland. Although his vanishing twin and chimera defence is clearly weak, the arguments put forward by Chris Campbell indicate that the test itself could be challenged for not being properly validated in either providing a false positive rate or for ensuring that the visual identification method was not subject to human error.

In Monday's San Jose Mercury News, Hamilton's lawyer Howard Jacobs was quoted as saying that they do plan to appeal to CAS. "I don't think it is futile," he said of the 2-1 decision. "It obviously wasn't a slam dunk for USADA."

Also see: Hamilton's Defence: In his own words