Twice Baked

by Rant on November 19, 2006 · 5 comments

in Doping in Sports, Floyd Landis, Tour de France

After taking a good look at Arnie Baker’s Slide Show 2.0 and going through everything that’s presented, a few things stick out in my mind. We’ve heard before about samples being mislabeled and goodness knows, LNDD’s admission of some “typing errors” has been hashed over quite a bit in the blogosphere over the last week.

Certainly, if you can’t be confident that the sample numbers match up with Floyd Landis’ bar-coded ID, then you can argue over whose results are being reported. And you can look at the CIR/IRMS data and argue over what constitutes a positive result. Can an elevated level on one metabolite be a positive result, or must all 4 metabolites be elevated in order to declare a test result positive? And then there’s getting into the nitty-gritty of the CIR/IRMS results and arguing over what particular peaks mean, and whether they truly indicate what the lab says they do.

And you can argue until the cows come home what a reasonable threshold value is for the T/E ratio in order to suspect an athlete is doping with testosterone.

But before any of that becomes part of the discussion, you need to make sure that you’ve tested a sample that’s not contaminated. And therein lies a problem with the data supplied in LNDD’s laboratory documentation package.

If Landis’ sample was contaminated, then testing should stop at that point because the values from further testing will be unreliable. So what is the standard for determining whether a urine sample being tested for testosterone doping is contaminated? Here’s what WADA Technical Document TD2004EAAS, quoted by Baker on slide 9 of his presentation, says:

The concentration of free testosterone and/or epitestosterone in the specimen is not to exceed 5% of the respective glucuroconjugates.

So, taking a look at LNDD’s own data for Landis’ B sample (as shown on slide 10 of Baker’s presentation), what do we find? The concentration of free epitestosterone was 7.7%, which by WADA’s own standards means the sample was contaminated. At that point testing should have stopped, according to the WADA protocol. No reliable conclusions can be reached from further testing due to contamination. In other words, the B sample can’t be used to validate the results of the A sample. Case closed. Whatever the truth might be, there can be no confirmation of the A results and Landis should be free to go.

But what about the A sample? There’s a few things wrong with that, as well. Before you get to arguing about the CIR/IRMS results, you need to know whether the T/E tests were reliable. And that’s another problem with the case. The T/E results show too much variation to be considered reliable. How can that be? Well, let’s try to keep this simple.

The original sample gets subdivided into “aliquots” for further testing. Now, if all of the hormones or metabolites are evenly dispersed in the original sample, testing the aliquots should yield results that fall within a certain range of the “average” reading. So, for instance, if the average reading is 5.0 and you have values ranging from 4.9 to 5.1, you can be fairly confident that the results are accurate.

But if the average is 5.0 and you have values that range from 1.0 to 11.0, how confident can you be that the true reading is 5.0? Not very. Why? Because the test itself is producing an extreme variation in results, so it’s hard to know which of the results is the “correct” one. So, let’s look at the results for the T/E ratios for Floyd Landis’ A sample. On slide 17 of Baker’s presentation, we see that one test on an aliquot of the A sample yielded a T/E ratio of 5.1, and another yielded a result of 11.4. How can that be?

Well, there are two possible explanations:

  • One, the testosterone and epitestosterone in the sample were not evenly dispersed in the liquid, or
  • Two, the test as performed on Landis’ sample was unreliable

In the first instance, you can’t draw any conclusions about the T/E ratio, as you don’t have a uniform dispersion and it’s not possible to determine the T/E ratio accurately. In the second instance, the test results themselves are unreliable. It doesn’t really matter which of these explanations you choose, the end result is the same. The test results can’t be relied upon.

So should a CIR/IRMS test have been performed at all on the A sample? No. There was no clear evidence that Landis’ T/E ratio had exceeded the threshold, because the testing for T/E was unreliable. Again, case closed. Or, at least, should have been. But we all know what happened next. The CIR/IRMS test was performed, and based on criteria that LNDD will not disclose, they determined there was an adverse analytical finding.

There’s a huge problem with LNDD’s criteria not being available. Not just for the Landis defense, but for every athlete ever tested by the lab. And that problem is that no one knows on what basis test results are determined to be positive or negative. Contrast that with both the American lab at UCLA and the Australian anti-doping lab. In Baker’s presentation, he’s able to discuss what their criteria are for a positive CIR/IRMS test. But LNDD will not release information about their testing standards. Why?

If they are performing tests using the same methodology and criteria as other anti-doping labs, they should not be afraid to say so. If they are using other criteria which no other lab uses, then there’s a problem. And here it is. According to the World Anti-Doping Code (shown on slide 25 of Baker’s presentation):

The purposes of the World Anti-Doping Program :

  • To protect the Athletes’ fundamental right to participate in doping-free sport and thus promote health, fairness, and equality for Athletes worldwide; and
  • To ensure harmonized, coordinated and effective anti-doping programs at the international and national level with respect to detection, deterrence and prevention of doping.

So what exactly are “harmonized, coordinated” anti-doping programs? That sounds to me like they mean consistent. Using similar, or the same, methods and standards to determine whether violations have occurred. By not making their lab standards accessible, LNDD gives the appearance that they have something to hide. Could it be that they don’t follow the same, or even similar, standards and methods for determining violations?

And if they are not following the accepted methods, are the methods they use accurate and reliable enough to determine whether an athlete has committed a doping offense? Judging by the fact that they continued testing on a sample that by WADA’s standards was contaminated, and they then “confirmed” the results of the A sample tests based on the results, it would be fairly easy to conclude that they are not following standard practices.

And again, for the analysis of the A sample, given the variation in the T/E ratios, one has to question the reliability of the tests or the abilities of person who performed those tests. In any event, it’s hard to draw the conclusion that the results are accurate and reliable.

In my mind, that should be enough for Landis to prevail during arbitration. But only time will tell. Right now, LNDD, the UCI and WADA are watching and waiting. I expect that when the arbitration hearings occur they’re going to come out, guns blazing. Whether they’ll be shooting blanks or bullets remains to be seen.

Debby November 20, 2006 at 12:11 pm

It seems to me, then, that athletes and teams are now in the position of deciding what races to partipate in based on the labs used by that race for testing. If it’s not a lab with testing criteria posted somewhere, then they should race elsewhere.

Someone once posed the idea that two samples (or one divided) should be tested in two different places, whether two labs or one lab and an independent medical staff hired by race organizers. That seems to be a much simpler solution to all this. If one sample is contaminated or handled improperly, the other sample could be used (of course, then who would break the tie…but I’m not going there).

Is there an international medical or scientific body, outside of sports, that sets standards for interpreting these tests? I’m just wondering why these labs have different test standards to begin with (though UCLA et al at least post their standards). WADA should re-write their rules to include, “all tests shall be interpreted according to the world standards and practices set out by the XXX agency…” But maybe that’s too easy.

Rant November 20, 2006 at 12:42 pm

Debby,

A truly independent body performing or setting the standards for the tests is an excellent idea. There are some logistical problems that would need to be worked out to ensure that samples are handled properly if more than one lab were involved. But if more than one lab was involved in the testing, then there would have to be consistent standards set for how tests are performed and so on. Whatever the case, the current situation with Floyd Landis certainly highlights the current problems with anti-doping testing.

– Rant

Mike November 20, 2006 at 6:26 pm

Rant, you have touched on the root of the entire case. This has gone so far beyond a simple lab test gone wrong, that I believe LNDD, and WADA are fighting for their very survival on this one. They sealed their collective fates when they dropped protocal and leaked the results. Now that they are facing a possible backlash of another “mistake” by the lab, they are backed into a corner. The facts available to the public thus far show the obvious, that the sample should never have been tested. Now we are to take it on faith that the “typos” are nothing to be concerned about, when in fact they are the very basis behind a valid lab test of any kind! Without a clear and concise chain of custody and control, no test can be considered valid. For some reason in this case that seems to be acceptable. Any reasoned person can only deduce that the LNDD and WADA must have an agenda of some sort. Why else not provide the information requested by Landis? Why not provide an acceptable answer for the admitted error? At least one beyond it’s no big deal. I can only deduce that LNDD and WADA feel more compelled to validate their existance, than create a fair and impartial system for the athletes. I fear that this will not be the last time an athlete is unjustly accused prior to “due process”.

Pops November 20, 2006 at 7:18 pm

I’ve been following the TBV discussions for a while, and one item that has occurred to me is why there is not a more visible partipation of the cyclists (as a group) in the testing. Do the cyclists not have significant medical supports that profile their blood during training? There shouldn’t be the level of shock and surprise that has occurred here… Could the competitors not hold “C” samples in escrow? Should there not be more transparency of the testing itself? I admit to being somewhat of a new convert to following cycling, but I truly think that the cyclists have the most to gain by having a self-policing and self-defence strategy. If someone is dirty, let the teams sort it out – if someone is clean, they shouldn’t be subject to this level of obfuscation and the predatory actions of the labs, WADA, etc.

Rant November 20, 2006 at 7:40 pm

Pops,

You raise some good points. I’ve seen a few other people make very similar suggestions. But I don’t know how common it is for the average pro to have medical supports profiling various hormone and other levels during training. Lance is said to have done such things, but he had the money to do so. The average pro doesn’t earn nearly as much and may not be able to afford to pay for such testing and other related expenses. Perhaps the teams should be the ones who pick up the bill.

There is a cautionary lesson in what’s happened to Landis, and that’s the idea that the riders or the teams should do just what you say, in order to counter the very abuses that have occurred in Floyd’s case. I don’t know how well self-policing would work in practice, but I’d certainly be willing to give it a chance. It’s bound to work better than the current system.

To be successful and credible, the testing system needs much more transparency than the current process affords the athletes. It also has to respect their rights to privacy before a case has been heard. The damage done to Floyd, no matter the outcome, may never be undone. If he were guilty, that’s not so bad, it’s the price you pay for cheating. But for an innocent person, it’s too high a price to pay. Justin Gatlin’s case didn’t come to light until it had progressed much further. It’s a shame the same can’t be said for the Landis case.

– Rant

Previous post:

Next post: