"So there is no doubt in my mind he (Lance Armstrong) took EPO during the '99 Tour."
Dr. Michael Ashenden began his career as an exercise physiologist with the Australian Institute of Sport. After assisting in the development of an EPO test for the Sydney Olympic Games, he left the AIS to focus on battling blood doping. In 2005, Dr. Ashenden was among of group of scientists who questioned the validity of a physiological
study on Lance Armstrong, a
dispute that led him to serve as an expert witness in an arbitration case involving Armstrong and a bonus payment for winning the Tour. Dr. Ashenden kindly agreed to speak with us and shed some new light on that controversy. He also helped us analyze the 6 positives from Armstrong's '99 Tour samples with a level of detail never before made public.
Personal background, the 2000 Olympics, EPO testing
Andy Shen: Can we start with a little background on you, starting with the Australian Institute of Sport? I guess at that time you were doing some work on blood doping, but you were also doing some work on performance enhancement.
Michael Ashenden: I was employed as an exercise physiologist with the AIS, and my job was to do physiological testing on the athletes, for example, to give the coaches feedback and assistance with their training programs. At the same time I was doing my PhD thesis that was studying what happened to athletes' blood when they were exposed to simulated altitude, hypoxic tents and such which have since become popular.
And that dovetailed nicely with research at the time that was looking for indirect markers of EPO, because not coincidentally I guess, the changes in the blood that we see with EPO were sometimes similar to changes you'd see at altitude, though dramatically on different scales. But it was easy for me to transfer my research and my knowledge there over to this EPO field and I guess that's where I began my career in anti doping, being part of the team in Australia who were working on what's since been called the Sydney Blood Model. And from there I left the institute and I've been working as a freelance researcher not just in EPO doping but other kinds of blood doping. Blood tranfusions, blood substitutes, pretty much any avenue I think an athlete might be tempted to abuse, we try to conduct research in those areas.
AS: The work you did for the Sydney Olympics, that was in 2000?
MA: Yeah, at that time there was no test at all for EPO, and the UCI was using a 50% hematocrit rule to stop athletes from competing, but it couldn't go any further than stopping them competing. So, there was a group in France that was researching what's since become known as the urine test for EPO, and we collaborated with them, we shared samples with them, to help the research move forward. But we took a different strategy, we looked for indirect evidence.
We were looking for changes in the blood that were not only apparent in the period when the athlete was using EPO, and that's the period when EPO is still in the urine, but blood remains disturbed for several weeks after you stop taking EPO as well. The urine test can't help you there, but the blood test still gives you a signature, depending on how much EPO you took, a couple of weeks after you stop injecting.
AS: The EPO test was implemented for the 2000 Olympics?
MA: At the Sydney Olympics they had the two part test. There was an initial blood screening, and if those results exceeded the threshold that were put in place, then the urine was analyzed. And the criteria back then was that you had to fail both the blood and the urine test in order to be found guilty of having used EPO.
Since that time WADA has revised the rule and now you only have to fail the urine test. Whether your blood fails criteria or not is not taken into account in today's test.
AS: So just to be clear, the urine test looks for the actual presence of synthetic EPO, but EPO leaves your system in two or three days?
MA: EPO is a hormone, it's a very small molecule, and it's present in very very low concentrations in the bloodstream, even lower concentrations in urine. And the half life of EPO is something in the realm of eight to twelve hours, so one day after you've had an injection the levels are dramatically lower. Usually three or four days after you've had an injection all traces of EPO have left the circulation or at least aren't present at a high enough level for the urine test to be a definitive piece of evidence that EPO is being used.
AS: When you test positive in a urine test, it's not a yes or no thing, it's a percentage and a threshold, is that correct?
MA: I think that's arguable. It's a test that discriminates, it puts in different positions on the gel, synthetic EPO and natural EPO. Now, there is no confusion when you see it on the gel, when there's synthetic EPO in the sample. It's simply in a different position to where the natural EPO occurs.
So when you say it's not yes or no, you can see visually if there is synthetic EPO in the gel. They build in some allowance, some tolerance, the positivity criteria that are in place today follows specific rules. And even though there's EPO in the gel, unless it fails those specific criteria a sanction isn't imposed.
AS: Ok, so in other words, the tester will know you're using it, but you won't be busted for it if you're below a certain number.
MA: Yeah, and there are situations like that floating about today. Where it's clearly a pattern that an athlete has been using synthetic EPO, but for different reasons the pattern doesn't follow a specific criteria. So the answer is, yes, obviously they're using EPO, but they weren't sanctioned because the samples didn't fail a very specific criteria that were applied.
AS: And this number is a percentage of isoforms?
MA: That was the first generation of tests. Since then the positivity criteria has been modified, and now it looks at several different aspects, not just the percentage of basic isoforms.
EPO use in 1999
AS: I want to go to the '99 Tour samples. Just to set the scene, in '99 there wasn't a test in place for EPO, and
Frankie Andreu told us there was no anxiety about using it, because as long as you made sure your hematocrit was below 50% you knew nothing could happen to you. Would it be fair to say that at that time it was pretty easy for cyclists to beat the test, or use EPO with impunity?
MA: Well, I wouldn't say that it was easy for them to beat the test, because there was no test. Simple as that. There was no way, before the Sydney Olympics in 2000, where an athlete could be found guilty of using EPO, because there was no test in play. There's no reason for an athlete to be careful using something for which they can't be caught.
AS: As long as they kept their hematocrit below 50.
MA: And that's a relatively easy thing to accomplish. I mean, you can either use careful adjustments of your dosage, or you can use saline to dilute the blood. It's a relatively simple and fast procedure to bring your hematocrit below 50.
AS: Yeah, in fact, from the time you're notified of a test you can actually dilute your blood fast enough to beat the test.
MA: Yeah, it was quite disturbing for me to be told that right up until perhaps 2004, the UCI weren't actually chaperoning riders between the finish line and doping control. So not only was there an opportunity for them to dilute their blood before a blood test in the morning, there was also a very real opportunity for them to manipulate or mask their urine before they provided their doping control sample. That wasn't important pre 2000 because there was no urine test for EPO, but after 2000 there was still, to me, unacceptably large loopholes for an athlete, even if they've been using EPO, to still escape detection. Particularly by masking their urine, in between the time they crossed the finish line and dope control.
'99 Tour urine samples re-tested in '05
AS: Let's go back to the '99 urine samples, these were B samples which were preserved. Was it for academic reasons that they re-tested, to get a sense of how things were at the time?
MA: I mentioned earlier there'd been revisions over time of what the positivity criteria were. Initially it was 80% basic isoforms. The research that was conducted with these samples was informing them of whether new criteria they were considering applying would have been effective in catching athletes in previous events.
The only kind of samples that are useful in that context are samples that have got EPO in them, 'cause then you could say by criteria A you'd fail, but by criteria B you didn't fail, and by criteria C we saw nothing at all. And that was the purpose of the Paris investigation - to go back, to look at samples, and to see how the different criteria applied. And it was, I don't think it was cynical, it was realistic, they realized that the most likely samples where they would find EPO were samples collected before the EPO test was introduced. And that was the '99 Tour de France.
Lance Armstrong's '99 samples test positive
AS: So out of the 87 usable samples that they gathered, they got 13 positives and 6 of them belonged to Lance Armstrong.
MA: Depending on which criteria you applied. Yes, six of them failed the definitive criteria. There were another two samples in fact where the EPO was visually there in the gel. You could see it was there, but for one reason or another, the percentage isoforms weren't calculated, or had to be re-analyzed, or it was a little bit too faint to get a definitive result. Yes, there were six samples with EPO in it, and there were another two samples where it was pretty plain to a trained observer that there was synthetic EPO in those as well.
AS: You were able to analyze the results, correct?
MA: I interpreted the results. They assessed each sample according the different criteria, and those were the results that we were given.
AS: I found it kinda interesting, we've talked before this, you found some very interesting things about those results that really were not widely publicized, the way the percentages fluctuated.
MA: One of the things, I guess there's been misinformation in this particular area - is that the samples weren't analyzed properly, that they were analyzed using a different protocol than what was used in proper dope controls - and that's just not correct. Obviously in research where the data you come up with is going to govern how you do testing in the future, you're exceptionally careful with these measurements. You want to make sure that you don't make any mistakes. And you want to make sure that you, for example, weren't looking at urine that has been contaminated with bacteria, or isn't what we call unstable urine, where sometimes the bands shift not because of EPO use, but because of some other factors. So all of these checks and cross checks were put in place with these samples, so the data is valid. The laboratory, I've checked with the people who did the analysis, and I very carefully went through it with them. They're absolutely 100% sure that these results are valid.
And as far as the fluctuations you speak of, when we took the samples' dates, and matched them with the percentage of isoforms, and overlaid that with the performances during the Tour de France, then a clear pattern begins to emerge. You can see that on some days there's a preponderance of EPO in the urine sample, perhaps on the next day they come down a little bit, then they come back up, which is suggesting you've taken another EPO injection.
You don't have EPO every single day. You might take it every two or three days. So your values go up or down according to when you took those injections and when those urine samples were taken. Now, you overlay all of those data together and you can begin to see a pattern that's consistent with EPO use.
Stage | Vial # | Visual Interpretation | % Isoforms | Stage description |
Prologue | 160297 | + | 100 |
|
1 | 157372 | + | 89.7 |
|
2-7 |
|
|
| Out of lead, not tested |
8 | 186584 | + | To be reanalyzed | Metz ITT |
Rest day |
|
|
|
|
9 | 185557 | + | 96.6 | Sestriere |
10 | 185479 | + | 88.7 | L'Alpe d'Huez |
11 | 185476 |
| Sample missing |
|
12 | 185475 | + | 95.2 |
|
13 | 185895 | + | Weak intensity, no % recorded |
|
14 | 186397 | + | 89.4 |
|
Rest day |
|
|
|
|
15-20 |
|
| Undetectable, insufficient EPO in urine |
|
AS: The prologue had the highest number, 100% if I remember correctly, and the next day it goes down a bit. Same thing before the first mountain stage, etc.
MA: And the unusual thing about the prologue sample is that the prologue was run in the morning, and the sample was collected at 9:45 am. Now, every other sample during the Tour de France is collected in the afternoon, after the stage is finished. This sample was collected very early in the morning, and there was 100% basic isoforms, which is saying that 100% of the EPO that was showing up in the gel was synthetic EPO. There was no endogenous EPO visual in the gel.
The possibility of tampering
AS: I guess I should set the background a little bit more now. This study was done for research purposes so the urine was marked with code numbers and there was no way for the testers to know who the samples belong to. It was only through some subterfuge by some French reporters that it was revealed that the six positives belonged to Armstrong.
MA: Well, again, there's been a lot of disinformation about this. The laboratory absolutely had no way of knowing athlete identity from the sample they're given. They have a number on them, but that's never linked to an athlete's name. The only group that had both the number and the athlete's name is the federation, in this case it was the UCI.
The UCI had those documents, and an investigative journalist, Damien Ressiot from l'Equipe, went to the UCI and said, "Can I have copies of Lance Armstrong's doping control forms from the '99 Tour?" Now, the UCI had to go to Lance Armstrong and ask his permission, which he gave them. Now, Lance Armstrong gave permission to the UCI to give these doping control forms to Damien Ressiot. Damien Ressiot took those forms, which have the athlete's name, obviously, and the sample number, so he matched the sample number with the results from the laboratory that had the sample number and the percentage of isoforms. And in that way he linked the percentage of isoforms with the number, the athlete's name, and in that way identified them as Lance Armstrong.
AS: Right. So the lab is carrying out these tests blindly, and you showed me this statistical study of the odds of them
tampering and successfully framing Armstrong, and it was 1 in 300.
MA: There was only two conceivable ways that synthetic EPO could've gotten into those samples. One, is that Lance Armstrong used EPO during the '99 Tour, and we've since found out that there were teammates from US Postal in that '99 Tour that have since admitted using EPO while riding for US Postal in that Tour.
The other way it could've got in the urine was if, as Lance Armstrong seems to believe, the laboratory spiked those samples. Now, that's an extraordinary claim, and there's never ever been any evidence the laboratory has ever spiked an athlete's sample, even during the Cold War, where you would've thought there was a real political motive to frame an athlete from a different country. There's never been any suggestion that it happened.
However, Lance Armstrong made that claim. Now, it's very easy to go back and assess the possibility of that scenario. We know the laboratory could not have known which samples belonged to Lance Armstrong. And we also know from the results, how many of Lance Armstrong's samples had EPO in them, and when during the race it occurred. Now the odds of the laboratory randomly selecting Lance Armstrong's samples out of those 87 samples, and let's just do it conservatively, just 6 times, 6 times they got his samples correct out of 87 possible tubes, the odds of that occurring are at least 1 in 300.
So we come back to the original scenario. Either Lance Armstrong used EPO during the Tour, or the laboratory spiked his samples, and we know the probabilty of that happening was at least 1 in 300.
(I needed to reassure myself that tampering was inconceivable, so I did some follow up with Dr. Ashenden. Click here if you're interested in what it would've taken to spike these samples.) An irrefutable profile
AS: And of course, if you take it to the next level, let's say, not only will they have to spike it, they have to spike it in a way that when positive samples are on concurrent days, the second day has to be a lower percentage. And not only that, when they spike the prologue sample they have to spike it really high because it was after a short effort and it was tested earlier in the day. Now if you take those factors into consideration the odds become astronomical, don't they?
MA: I honestly can't conceive how you could possibly do that. I don't understand how you could inject enough EPO so that the percentage was slightly lower on the next day, it just beggars belief that you could adjust the amount of EPO you put in a sample by such a miniscule amount. And to be quite frank, it doesn't hold up to scientific scrutiny, it's a fantastic claim in the literal sense of the word, it's not backed up by a shred of evidence at all, and I think it needs to be taken on that merit.
AS: So outside of deliberate tampering, is there any way contamination, degradation, is there any way synthetic EPO appears in urine because of contamination, degradation, bad handling, bad refrigeration, anything?
MA: The short answer is no but I have to clarify that. There is evidence that sometimes if a urine sample is stored unfrozen, there can be some contamination of the sample that shifts the band up towards the area we associate with synthetic EPO. Now, it can still be distinguished but it makes it more difficult. There is a test and this is in place throughout laboratories today, they can determine whether or not the sample has that unstable profile.
That possibility was excluded in all of these samples. So yes, it's conceivable that contamination can shift the band, but it didn't happen in this case, that was definitively excluded. There is no way that synthetic EPO can suddenly appear. It can disappear, you could conceivably have degradation where synthetic EPO could break down, it's not likely but it's conceivable. But in that scenario you've got synthetic EPO disappearing, not appearing. It's breaking a pretty fundamental law of physics to say you can generate a molecule of EPO from nothing.
AS: So based on that, you can definitively say that Lance Armstrong used EPO in the '99 Tour. No doubt in your mind.
MA: There is no doubt in my mind these samples contain synthetic EPO, they belong to Lance Armstrong, and there's no conceivable way that I can see that a lab could've spiked them in a way that the data has presented itself. So there is no doubt in my mind he took EPO during the '99 Tour.
The rest of the '99 samples
AS: The other thing that struck me about these results, which I was surprised never came up before, was that if you take away those 6 positives, you have 7 remaining positives out of 81 samples. That's 8.6%. Does that say to you that at that time the peloton was relatively clean?
MA: Yeah, it's an interesting observation, 'cause you cast back to the '98 Tour, obviously it was a debacle. And, I've heard anecdotal or off the cuff remarks, that '99 was a new beginning. It had gotten as bad as it could possibly get, or so we would've thought, and '99 was, "Ok, let's start again, we've really got to make an effort to be clean this year."
Well, obviously, based on Lance Armstrong's results, he wasn't racing clean. But for the rest of the samples collected during the Tour, relatively speaking there wasn't a very high prevalence of EPO use in the rest of the peloton, at least in the peloton that was tested, which was your top 3 place getters, for example.
The prologue was interesting. First race of the event, every one of those samples had EPO in them. So it seems a little odd, the first day of the next year's race, and all of your place getters have got EPO in their urine. On the one hand, yes, it seemed less prevalent than you would've otherwise thought, but on the other there's still evidence there was doping in the peloton. Not just by Lance Armstrong.
AS: I guess it's possible that some guys were injecting during the Tour, and some had an EPO program leading up to the Tour and counted on the effects to remain with them?
MA: It's conceivable. It's widely known that you don't have to be using EPO to get the benefits. You can have a treatment regime that could last as little as ten days, and the benefits are substantial and they'll stay with you for four weeks afterwards. And certainly for the Tour, which is three weeks. So, you don't have to use EPO during the Tour to get the benefits.
AS: I just want to go back to the percentage. Obviously, stage winners are always tested, and there were, I believe Cipollini won four, Steels won three, Etxebarria won two, so, not that I'm accusing them, but there's a chance that some of these positives are from the same person, so there's a chance that the number of people positive is even lower than 8%. And not only that, a great deal of these samples are from stage winners, so they're the stronger riders. So the samples are a skewed sampling of the entire peloton.
MA: Yeah, that's correct.
AS: So you could say as a whole it might've been 8% or less.
MA: And there is no way to identify who those other samples belong to without getting access to the medical records and matching the numbers to their names. So, I've read reports, but I've never seen documented evidence to link names to numbers.
AS: It's interesting because when I spoke with
Paul Kimmage he made a pretty big deal about that year's Tour, that it was supposed to be this Tour of redemption, and his point was that Armstrong came along and brought things back to the old ways. These results lend that belief a little bit of credence, don't they?
MA: I think there's a couple of things that strike me as well. Yes, these results are consistent with that argument. The other is that we know how Armstrong performed before the '99 Tour. '93, '94, '95, '96. And look, a couple of those races he couldn't even finish, another race I think he's an hour and a half behind. Specifically in the time trial he was dropping minutes to the other competitors.
Now compare that with '99, and it's a helluva transformation. Instead of dropping off and not being competitive, he was actually dropping the rest of the peloton off. So something dramatically changed in relation to Armstrong versus the rest of the peloton across that period of time. That's unarguable.
There's, as we've been talking about, pretty unequivocal evidence, well, it is unequivocal evidence, that he was using EPO during the '99 Tour. Now, that would go a long way to explaining that reversal in competitiveness in Armstrong v. the peloton.
...for the rest of the interview...hit this link
