Tuesday, April 24, 2007

What the?

The AFP, via Turkish Press, rehashes L'Equipe under the headline "Landis Ultra-positive", with a body that refutes that with a few things we haven't seen anywhere before:

L'Equipe reported that varying levels of synthetic testosterone have been found in urine samples tested this week
the results from the French laboratory involved in testing seven of Landis's samples revealed that "several" of those samples contained synthetic testosterone, albeit at lower levels than his original positive test.

and then finally (emphasis added):
several of Landis's samples from last year's Tour contained synthetic testosterone, albeit at levels that were not sufficient to warrant a "positive" test.

If we read this correctly, there aren't positive test results in the newly tested B samples -- only claims that there is exogenous testosterone present, which are unsupportable if they aren't positive.

If this information is true, it casts the leak and the reporting of the leak in an entirely different light.

We await clarification and confirmation.


Overnight, consensus seems to be that the AFP story has confused the T/E screening tests with the IRMS CIRs, and that there are no new facts -- the tests are not known to be "Ultra positive" based on the data available, nor are they known to be indicative of exo-t but not at levels to be "positive".

This hasn't stopped the headline or the story from racing around the world, following timezones.


(nb: TurkishPress often runs wire stories before anyone else, possibly because of timezones).


swimyouidiot said...

I don't understand what this AFP article is claiming. It sounds like they have confused and conflated the screening test and the IRMS test. They don't test for "levels" of synthetic testosterone, just presence or absence. There is no positive "level" for exo-t. Are they trying to say that the 12C/13C ratio wasn't as high as in the stage 17 test? What does it mean to say that the samples "contained synthetic testosterone, albeit at levels that were not sufficient to warrant a 'positive' test"??

There seems to be something I'm missing here and not stating properly, but it's late and I'm not thinking to well right now. Can someone state this properly?

tbv@trustbut.com said...

My reading suspicions are that (a) someone is saying that negative values represent a "level" of exogenous testoserone; (b) the tested samples had "some" exo-t by that measure; (c) but not enough to be a "formal" positive. (d) but it's there! It's there! He's a witch!


marc said...

I think swim has the right interpretation. It's possible AFP has another source, but since the article that's linked to uses a translation of l'Equipe's headline, I think this is just a confused story about the original l'Equipe article. The article mentions that the seven additional samples were below the positive threshold--but means on the screening test. I agree with swim that the AFP or Turkish author ran the two tests together.


Thomas A. Fine said...

This is perfectly in line with the whole "traces" problem in the original L'Equipe article. But I'm starting to think that instead of this being a case of a journalistic "telephone" game, that maybe the distortions are coming directly from the leak.

If the person leaking the results doesn't really understand how the test works, this is exactly the sort of things they'd be saying with the typical numbers you find in these tests.

I started a topic more or less on this subject on DPF.


Death Bredon said...

If the illegal tests a LLND are not legally "positive" for exo-T, then the case against Floyd is over right?

zarghev said...

The positive have already been found: it is the A sample, and the B sample of last year.

The USADA probably only assess whether the process of the laboratory is correct. If it is, then Floyd Landis must have used some way to reduce his testosteone ratio, and the tests are necessarily a little diluted. However imagine doing the following experiment: take 50 samples, 7 from Landis and 43 at random from others. Test them blindly. If after testing it turns out that the top 7 are from Landis you have a strong suspicion.

Moreover, the levels required for becoming positive for one test, are much higher than the ones for reasonable suspicions. But if you have 7 times more than strange results (after the proof of the initial A/B sample), then you can even make calculation starting from which repeated suspicion becomes statistical "positive".

Pearl said...

I tried to post a version of this comment yesterday, but it did not make it through (at least anywhere I can see). Forgive me if I repeat for those who can see yesterday's comments on TBV. I have had a question from the beginning of this whole mess that I have never seen answered, and in light of the new "positive" IRMS tests, I think it is an even more important question. It is well known that FL had a TUE for exogenous cortisone injections into his bad hip. Cortisol and testosterone are similar molecules with possibly some common metabolic pathways. Is it possible that some of the carbons from exogenous cortisol could become incorporated into the testosterone molecules made by a cortisol-treated male's own body? Has anything been done to demonstrate that this does not happen? Has anyone ever done the IRMS testing for "exogenous" testosterone in the urine of any guys that have had cortisol (but not testosterone) injections? If so, I have not found any reports in the literature. Can anyone direct me to reports of such studies? If such studies have NOT been done, I don't know that the possibility that the cortisol injections caused some "abnormal" results on the IRMS test for "exogenous" testosterone can be ruled out. It needs to be, especially now that other "abnormal" IRMS results have been found, because IF the "abnormal" results were indeed due to the cortisol, then IT WOULD BE EXPECTED that more than one sample would give such results. IS IT POSSIBLE that FL's admitted alcohol consumption (or some other factor)led to a positive T/E result that caused the discovery of what looked like possible exogenous testosterone in his urine, but which was actually a false positive result due to the cortisol, and might that not also explain why the "abnormal" results are not clear-cut positive results? Unless there is some evidence that this is not possible, then I don't think anyone involved with the development or application of the IRMS test for exogenous testosterone in urine can walk away from this case with a clear conscience. Unfortunately there are individuals at WADA, USADA, LNDD, etc. that don't seemed to be troubled much by issues of conscience. If the question I have posed is not answered (and maybe someone reading this can answer it for me), I will never be able to accept that FL is guilty unless he someday admits he is. And, based on the actions of WADA, USADA, LNDD, etc. and their apparent disinterest in learning the truth rather than just getting a conviction, I will now not be able to believe the results of any drug testing on any athlete, positive or negative. Unless some serious changes are made in the process of testing and prosecution of positive results. If this whole debacle could somehow lead to much needed improvement of the whole drug-testing process in athletes, then at least some good would have come from it. But I'll not hold my breath.

tbv@trustbut.com said...

Wow, Pearl, I had to open a whole new window on another display to be able to answer!

First, most of the good public scientific discussion has taken place at the Daily Peloton Forums,, There are people there of all predilections and no consensus has emerged on the ultimate facts of the case.

Yes, the cortisol metabolites have been publicliy looked at. Most believe there is no metabolic path to directly affect the chemistry. There is suspicion it may be relevant to interference in the testing, called "co-elution", which would be resolved by examination of the mass-spectrometry data files of the tests, which have not yet been provided to Landis.

It is known that Alcohol affects the T/E; whether it affects it enough to skew Landis's results is unclear, but there are some other oddities in the way that was done that may make the results there unreliable.

It is speculated that Alcohol -may- affect the CIR, but this is not backed up by any studies. Tom Fine at DPF and comments here is a champion of this approach, and he's mostly dismissed by the DPF scientists who think Landis is guilty.

It appears to us the basic attitude of the doping system is: they're all guilty, what's your problem with our methods?


Sacky said...

I've been away from scouring the Floyd-blogs for a few weeks, but this recent news has dragged me back.

I haven't seen any other mention of it, so I'll bring it up (and if it has been mentioned, hopefully someone will point me in the right direction...)

Anyhow, I recall a lot of discussion about the isotope testing. In particular, I recall discussion regarding the ramifications of the isotope test incorrectly detecting the legal cortisone injections for Floyd's hip and the lab reading that as synthetic testosterone.

I don't recall how far this discussion went or if it ever bore any fruit, but the stuff I read led met to believe that it was entirely possible. Whether through mis-calibration or purely natural means, the arguments I read convinced me that it is quite possible that the "synthetic testosterone" we all keep hearing about is really the medicine Floyd was legally allowed to take for his hip.

And how does this affect the recent news?

To me it makes perfect sense. If the argument is that the test is not positive and some other substance was detected, it makes perfect sense that the isotope tests run on EVERY OTHER SAMPLE would come back with similar results.

And while I'm reaching for possiblities here, I'll suggest that it's possible that after suffering through Stage 16 Floyd may have upped his hip medication. If this is what the IR tests are detecting, perhaps an increase in the regular dosage is what initially set off the Positive A test on Stage 17...

Does any of that make sense? I hope so.

swimyouidiot said...

Pearl, just to save you a little time searching, at dailypelotonforums.com go to the Floyd Landis Case area, and find the "Mass Discrimination in GC MS C IRMS" thread and start reading at post #198. There are several posts about the cortisone issue.

I'm a science "dummie," and I am basically repeating what TBV already said, but I think what it boils down to is that there are no studies which show that cortisone could break down into the same metabolites as testosterone. However, as TBV said above, one of the metabolites of cortisone could conceivably have been measured along with one of the metabolites of testosterone and thrown off the tests (all the tests). The testing data should be able to show whether this happened or not, but LNDD (or USADA) has not produced that data. Some of us suspect they don't have the data (at least for stage 17), but that's only (informed) conjecture.

Read the DPF stuff and you'll see the arguments.