Platinum chemo and the EPO test

[garnetstar]

So, I'm still thinking about how scientifically valid the new EPO test can claim to be, and there's another problem with it that may, in fact, be unique to the 1999 Tour de France situation.


That is, that Lance Armstrong was exposed to very high doses of the chemotherapy drug cisplatin, which is actually platinum. For all I know, some other cyclists may have been exposed to heavy metals, through chemo or in some other way, so the EPO testers should take this into account (although, most people who are exposed to high doses of heavy metals are dead, so this may not be a huge problem).


I worked in the lab in where we first discovered how platinum binds to cancer cells and causes them to die, so I'm kind of interested in this.


Platinum, and all heavy metals (like, you know, mercury and lead and thallium and, to some extent, arsenic), are so poisonous because they can't be gotten out of the body after exposure. There's no medical treatment that's very effective in removing them. Also, while the body itself does try to excrete them (aka "clear them"), this natural process is so slow and inefficient that, really, you don't have a good chance of clearing enough metal to live long.


I had a friend who was exposed to mercury (not even that much, either!) in a lab accident. She was decontaminated right away, and they gave her every known medical treatment to try to remove the mercury. But, she soon went into a coma, and after ten months she died.


At autopsy, they found that, after ten months of her body trying to clear the mercury, she still had 71 times the fatal dose of it in her body. That's how slow the body is at gettiing rid of heavy metals.


Platinum is very, very toxic, just like mercury. That's how it kills cancer--it doesn't discriminate, it kills both healthy and diseased cells in your body. It's just that the cancer cells are killed more quickly. During chemo, you receive doses of platninum that are as high as possible, while still under the level that would kill you along with the cancer. That's why platinum chemotherapy is so agonizing--it is literally slow poisoning.


(BTW, don't let this stop any of you guys from getting platinum chemo *pronto* if you get testicular cancer, or any of the many other kinds of cancer that platinum is effective against. Testicular cancer used to be 100% fatal--with platinum chemo, it's now close to 100% curable. But yes, it is unpleasant.)


When you're exposed to platinum, it accumulates in the kidneys, sometimes causing kidney damage, in fact. (I read somewhere that Lance Armstrong actually sustained some slight kidney damage, but this is unconfirmed.) The point is, the way that the body tries to clear platinum is by sending it to the kidneys, *so that it is excreted in urine*.


In the kidneys, platinum binds to proteins and anything else that's being excreted in urine, and appears in the urine bound to the proteins that are excreted. (I mean, it doesn't appear in the urine as just free-floating platinum. It's always chemically bound to some protein, and is excreted when the protein is.) Any protein that wasn't bound to platinum when it was in your blood, will get bound to it when the protein passes through the kidneys, since there's so much platinum accumulated there.


Therefore, Lance Armstrong, and all other platinum chemo patients, will be excreting protein-bound platinum in their urine for some time after the chemo ends. It should be for a *very* long time, considering how long it takes for heavy metals to be cleared from the body. No one has ever studied how long platinum is excreted in the urine of former chemo patients, because no one has ever had any reason to care (before now).


Naturally-occuring EPO is a protein that is excreted in everyone's urine, and some of it should have platinum bound to it in the urine of former chemo patients.


Now, the new EPO test is based on the fact that there are slight chemical differences between natural and artificial (i.e., injected) EPO (see that New York Times article that I posted a link to in the other thread). The new test separates the natural from the artificial on that basis, that they are chemically different. Anything that is is slightly chemically different from natural EPO will be called a positive for injected EPO in this test.


But, having platinum bound to EPO makes the EPO chemically different (*way* different) from natural EPO, and so it will show up as a different kind of EPO than natural EPO that has no platinum bound to it. In other words, it will show up as a false positive for injected EPO, a false positive for doping. Former platinum-chemo patients (and anyone who's been exposed to high-dose heavy metals) should show false positives for artifical EPO for some time after their last dose of platinum.


I've read that Lance Armstrong received his last dose of platinum on December 13, 1996. It is more than probable, actually rather likely, that his urine still contained detectable levels of platinum-bound EPO in July 1999, only 2 1/2 years later.


So, if an unambiguous positive test for artificial EPO showed up in Armstrong's urine from the 99 Tour, the scientists have to rule out that it's not actually platinum-bound EPO, instead of artificially injected stuff.

(Really, they should do this for all the cyclists, since some of them might have had high-dose heavy metal exposure to something, not just platinum, at some ttime in their lives. But, Lance Armstrong seems to be the only cyclist that is known for sure to have been recently so exposed.)


This isn't going to help determine whether Armstrong doped or not in 1999: by now, nine years after his last exposure to platinum, there probably aren't detectable levels of platinum being excreted in his urine anymore. The body *does* slowly clear heavy metals, and after nine years---well, it has had a long time to do so.


The only way to find out would be to get some EPO from Armstrong's 1999 urine samples, and test it for the presence of platinum. But, I gather that there are no more of his urine samples to test, that they've been used up. (Or else, try to persuade some recent platinum chemo patients to donate their urine samples, and see if platinum shows up bound to their natural EPO, and if that gives a false positive in the new EPO test. Good luck getting funding for this!)


The scientists never thought of testing for platinum, because only one cyclist in the 99 Tour seems to have been so exposed. It's hardly worth the cost (millions) of an entirely new scientific study to find out if other chemo patients excrete platinum-bound EPO, just to test only one cyclist!


And, body chemistry differs from person to person: Armstrong's platinum-bound EPO signals in the test may differ from those of other chemo patients. Since there is no more of his urine to compare to the other pateints' results, this can never be determined.


I'm not saying that Armstrong didn't dope in 1999, or that his platinum chemotherapy would for sure give rise to false positives in his urine samples. It's just that, before these scientists say that he (or anyone exposed to heavy metals) has *scientifically* tested positive for EPO doping, they need to rule out this rather-likely possibility.


But, if the lab doing the testing had not been so careless as to let their data be leaked (which is almost inexcusable in reputable scientists), and therefore have Armstrong's tests publically identified as "positives for injected EPO", this issue would never even have arisen!


Because, as I say, so few people are exposed to heavy metals (and live to tell about it) that there'd be no reason to worry about this possbility in large-scale screening of cyclists in future Tours. I mean, how many times would it ever even come up as a possibility?


If this lab hadn't been so careless with their data (and, I hate to even suggest that they deliberately leaked the data, because this is *completely* inexcusable in scientists, and they'd get expelled from science if it was true), then this possible problem with their new test wouldn't ever have arisen, and so needed to be ruled out.

But, they*were* careless, so now they must address the issue before they call their results on Armstrong scientific or reliable. (You know, that's what I'm really ****** about in this whole affair, that the sceintists leaked their data! I don't really care if any cyclists dope or not! (sorry).

 

[VeloFlash]

You have two problems with your hypothesis.

1. If 6 samples in 1999 were false positive through your hypothesis (platinum remaining in the system for many years after cancer treatment), why were not the other 9 samples (11 if you believe LA) also (false) positive?

2. Why were not the urine samples taken 2001 - 2005 also false positive?

The logical argument to 1. is the French urine test only detected the presence of EPO use within 24-36 hours of administration and the other 9 samples were outside this window. Your platinum hypothesis would have every sample test scoring a positive.

In response to 2., EPO was being continued to be used by the peloton when tests were introduced. But they discovered methods (microdosing, overhydration) to avoid detection. The methods would not have avoided the platinum presence in urine making naturally occurring EPO "different" and being analysed as a positive.

 

[garnetstar]

Veloflash,

Thanks for your thoughts! I really like discussing this stuff, especially with someone who knows about cycling, since I don't at all.

(I'm going to be quoting a lot in this post from the EPO testers' Nature paper, so here's the link to it: http://www.nature.com/nature/journal/v405/n6787/index.html. The article is on page 635.

Anyone can download it, for free if you're subscribed to the database, or for a small fee to buy that individual paper.)




VeloFlash: "You have two problems with your hypothesis."

I think that you may be a little mistaken about what my hypthesis is. (Actually, it isn't a hypothesis, it's a criticism of the EPO testers. Scientists just love to criticize each other's work, with good reason, too.)

As I said, "I'm not saying that Armstrong didn't dope in 1999, or that his platinum chemotherapy would for sure give rise to false positives in his urine samples." I'm only saying that the latter is a *likely* possibility, one that has to be ruled out before someone publically announces that a cyclist who's had recent heavy metal exposure has scientifically tested positive for doping.




"1. If 6 samples in 1999 were false positive through your hypothesis (platinum remaining in the system for many years after cancer treatment), why were not the other 9 samples (11 if you believe LA) also (false) positive?....The logical argument to 1. is the French urine test only detected the presence of EPO use within 24-36 hours of administration and the other 9 samples were outside this window...."



"the French urine test only detected the presence of EPO use within 24-36 hours of administration and the other 9 samples were outside this window."

Absolutely true, and that well might be what really happened. But, unless the scientists publish their data in a peer-reviewed journal, including controls for possible heavy-metal effects, no one will ever know.



The scientific objection that I have is that these results could undoubtedly also have arisen from the new EPO test's central problem--that most of the results it gives seem to be ambiguous, and cannot be called positive or negative with any certainty (see my other thread on that, and the NYT article link).

So, in the published study of the 1998 Tour samples (the link above), only 2 of the 14 samples tested by the new test look like they gave completely unambiguous results. As I said, any samples' tests that were too ambiguous to call were undoubtedly tossed out by that paper's reviewers. (And, I'd like to point out again: in that published study, they only tested 14 of the 102 urine samples with the new test. Not even a statistically significant fraction.)

What good scientists are *supposed* to do is make *all* their test results available online in a link that's published right at the end of their paper, called "Supplemental Data". Then, if there wasn't room to publish everything in the actual paper, anyone who wants to can download the results of *all* the 14 tests from that Supplemental Data link. Then the results can be looked over to see which ones were ambiguous, and how many of the 14 were ambiguous, and how ambiguous they were.

We'd all be able to see for ourselves just how reliable this new test is. That is *standard operating procedure* when you publish a paper. That these scientists didn't do this is perhaps just carelessness, but it doesn't help their case any. BTW, I don't want to boast or anything, but the last paper I published had 12 MB of Supplemental Data made available for anyone to look over. Like I say, it's SOP.

The scientists say in the Nature paper: "The presence of exogenous hormone (artifical EPO) was always evident." We need to see the data, please, so that we can judge for ourselves. If you want me to think that your 12 unpublished tests from 1998 were unambiguous, prove it: show me the data. That is how science works, everybody should doubt your claims until you *prove* them beyond---well, a reasonable doubt, sometimes beyond any doubt.


Since this EPO test has some inherent ambiguity, it's likely that not all of Lance Armstrong's or anyone else's 1999 urine samples gave unambiguous results, in which there would be clearly two different EPO signals, one from natural EPO and the other from artificial (or, perhaps from platinum-bound) EPO.


If Armstrong gave 15 or 17 urine samples in 99, and they were all tested (do we even know if they were? They need to publish this!), the reason that the results of 9 or 11 of them haven't been revealed is almost certainly that those results were too ambiguous to call as positive or negative. (I mean, if they just got negative results because the injected EPO was outside the time-detection limit, why haven't they just said so? Just said "These tests were negative. However, our tests have a time-detection limit, so these negatives aren't positive proof of innocence.")


The lab has been suspiciously silent about what the results of these 9-11 tests were. That seems like a classic example of the very common phenomenon of scientists not wanting their pet theories (i.e., that their EPO test is near-100% reliable) to be criticized. And, as always in science, the burden of proof that that's not the case is on them.

Like I say, other scientists *really need* to see and review the data from all these tests, so that the ambiguous results can be sorted out from the unambiguous ones. Otherwise, no one will ever know what the testers are calling "positive" results, what they're calling "negative", and what they decided was too ambiguous to call. And so, we'll never know who really tested positive in the 1999 Tour.


In the Nature paper, the EPO testers very correctly and significantly conclude: "Our method for detecting recent exposure to recombinant EPO in athletes *could be useful* for in-competition controls... " They don't say "Our test is *conclusive and reliable proof of doping*" (as you've pointed out, it isn't proof of not-doping). They say it *could* be useful. Yes, it could, when they've refined it more and it gives fewer ambiguous results.



I'm going to try to post, in another thread or somewhere, the scientists' actual data that were published in their Nature paper. Because, that'll make it really clear what I mean here, and I'm real steamed up about this issue of accurate science (you may have already noticed that).




"2. Why were not the urine samples taken 2001 - 2005 also false positive?..."


*This* is exactly the point, that hits the nail right on the head.

If the lab had just done what they should have, not allowed public leaks and not identified a known heavy-metal-exposed cyclist as having tested positive, this issue would never have arisen. No one would care if platinum (Pt) causes false positives, no one would need to know, and no one would ever have had to spend any money or time finding out.

They should have published the 1999 data in a journal, and just reported "These were the number of positive tests we got", without identifying whose samples they were. Then, who'd know or even care if Armstrong's samples were false positives because of Pt? The heavy-metal situation is really unlikely to ever happen again, so no other large-scale EPO testing data is going to be affected by a few false positives that may have occured in just one study. And the overall reliability and usefulness of the EPO test as a doping control isn't going to be affected by a few false results, again, just in this one study. So what if the 1999 tests showed a few false positives (for any reason)? Nearly all tests for any drug sometimes do.


If they'd just acted like scientists should, I wouldn't have to be writing all this. And they wouldn't be ethically obliged to do all the extra work that I'm about to assign to them.


OK, here's what I'd tell them that they must do if they want Armstrong's results to be accepted as scientifically proven positives for doping. If they don't do these experiments, as a reviewer I would reject their paper as being unpublishable in any scientific journal:


1. Find out if Pt binds to EPO. Buy some natural EPO *and* some artificial EPO, and determine 1) if Pt binds to it, 2) how much Pt binds, 3) if Pt binds differently to natural vs artificial protein, 4) how strongly Pt binds and how long it stays bound, 5) if, in the physiological conditions present in the kidneys, does Pt bind to EPO, how much, natural vs artifical, how strongly, etc., etc. All these things can be determined with hard numbers, so do it.


2) Find out *how long* chemo patients excrete *detectable* Pt in their urine. It's already known that Pt is excreted in urine: doctors measure how much is in the urine during the patient's chemo, so that they know when enough Pt has been excreted, and so when they can give the patient another dose without killing them.

So: how long after chemo can you detect Pt? I'd bet on a few years, but that's nothing but an*educated guess*! Not scientifically sound. Every metal is a little different. Find out about Pt!

Screen former chemo patients' urine samples for Pt, from right after their last dose until no more Pt can be detected (for years, if necessary). Then plot curves of the amount of Pt detected vs time, fit the curves to an appropriate mathmatical expression (it's probably an exponential), and use that to extrapolate for other individuals.

Screen *large* numbers of patients, please (not 14--hundreds), because individuals' metabolisms vary, and we need to know a good lower time-detection limit and a good upper time-detection limit, during which we'd expect to see Pt excretion in different people.


3. Once you have all these urine samples from chemo patients with Pt in their urine, *isolate the EPO from those samples*, just like you do for your EPO doping test. Then *find out how much Pt in the original urine sample is bound to the EPO*. Pt, like all metals, can bind to all kind of things--cells, all kinds of proteins, small molecules that are present, etc. So, when you do isolate the EPO, how much Pt that was in the urine is bound to it, rather than to all kinds of other stuff? This amount can be exactly quantified, so do it.


4. Now, do your EPO test on all these samples. Test *all* of the following: store-bought natural EPO, store-bought artifical EPO, natural EPO that was isolated from a known-to-be-clean person, EPO that was isolated from a known-EPO-doped person, store-bought natural EPO that has Pt on it, store-bought artificial EPO that has Pt on it, and all the hundreds or thousands of samples of EPO that you collected and isolated over many years from chemo patients, from right after their last dose until whenever you can't detect any more Pt in their EPO.

(There are actually a lot more possibilities that should be tested for than I've given! But, I'm not going into them all.)


You will then, and only then, know what effect Pt binding has on the results obtained by your EPO doping test, and what effect Pt *chemo* has on the results, and, most important, *how long* Pt chemo causes any effect on your doping-test results.



The 2001 - 2005 Armstrong EPO tests were called as negatives.
(Is this new, 2000 test the one that they used during those years? Or the older one? They shouldn't be using a test that they are still doing active research on to improve, but, hey, that's not my problem.)

Well:
--Do people still excrete *detectable* platinated EPO after 5 - 9 years? What's the Pt excretion time window?
--What's the Pt level?
--If it's really low, can it be distinguished from natural EPO?
--What kinds of test results does this Pt level give (always positive? ambiguous? negative? sometimes one thing, sometimes another? anything at all?)
--Will this level of platinated EPO give a strong enough "false positive" signal to be counted as real?
--Or will the signal be so faint that they'll discard it as not true, and call it negative?
--Is 2 1/2 years very close to the time-detection limit for platinated EPO, and so by 2001, Armstrong's levels were undetectable?
--In that case, one might expect that the six 1999 "positives" were near the detection limit of the EPO test? Are they?
--Or are those "positives" very stong, clear, unmistakable signals? Could we *see* the damn data, please?

I could go on, there are a thousand more questions that need to be answered before we can *know for sure* the answer to your very reasonable question!



"Your platinum hypothesis would have every sample test scoring a positive..."

See, that's not true until we know the Pt time-detection limit, as well as a lot of other things.


Actually, they've got to establish first of all that any Pt can even be detected, not in the urine, *but in the urine-excreted isolated EPO*, of chemo patients. Their urine might be full of Pt, but not enough of it may exist bound to the EPO, and it is only the EPO that the doping test looks at. So, establish this first of all.


Then, how long after the last Pt dose can Pt be detected in urine-isolated EPO? A year? Five? Ten? How long should we expect to see Pt?

If Pt is excreted in detectable quantities bound to EPO, at first there will be a lot of Pt detected, then, as time passes, it will drop down (I think exponentially), and at some point, although Pt may still be bound to EPO, it'll be under the detection limit for Pt of any method that we have. So, while heavy metals may perhaps linger in the body forever, the amount being excreted in urine won't be detectable after a certain time.


So, no. Not every EPO sample from a once-platinated person would give a false positive unless you have *established by conclusive experiments* that Pt is still detectable in the EPO of urine isolated from chemo patients after the same time period from their last dose (for at least 2 1/2 years after their last dose in this case, which was Armstrong's point in time during 1999).



And, it gets worse: detecting that EPO still has Pt on it is one thing. Seeing what effect even low levels of Pt have on the EPO test results is quite another!

Does Pt make the test give unambigous false positives? How often--everytime, sometimes? (Please determine the statistics of this.) How much Pt has to be present to do that? Does Pt make the test give *ambiguous* results, that can't be called? How much Pt has to be present to do that, and how many times does it make the results ambiguous (please give the statistics on this, too).

Can you even distinguish, using the test, a sample that truly has artificial EPO in it from one that has platinated EPO? How often can you do this? How clearly? (Can you even distinguish platinated EPO from natural EPO, for that matter?) Show statistics and calculations of error on all this, please.

And, of course, what effect does time-since-last-Pt-exposure have on all these results? What happens over time? Do things change, and how?


5. Now, compare all this data to Armstrong's test results from the 1999 Tour--all of the 15 to 17 samples, not just the ones called as positives. Oh, of course, use exactly the same experimental procedures in gathering all the Pt data over all these years that you used in testing his 1999 Tour samples! Don't let the pH of the water change a little, like, from your water pipes getting more rusty, or let the bottles of chemicals get a little older, or let your EPO isolation procedure change at all, or anything! or your results won't be valid.

And, also of course, don't run *any* of the tests that I require until all your platinated samples are exactly the same age as Armstrong's 1999 samples, and have been stored in exactly the same conditions all that time! Just wait another six years, no problem.


So, once you do all that work (phew!), you can then scientifically prove that any "positives" (I mean, not "ambiguous" results that arise from the test's inherent ambiguity--see above) from Armstrong's 1999 Tour samples were *not* caused by his exposure to Pt. (Or, maybe you'll be able to prove that they were caused by Pt exposure, you'd have to do all the work to see if that's possible.)


*Then*, and only then, can you say that this likely possibility (Pt exposure) for false positives has been ruled out. And, you'll be able to say that your results must therefore show that Armstrong was doping in 1999, *if* the positives on his samples were unambiguous enough to unequivocally make this call. (A very big IF--you'd have to prove this by publishing the 1999 test results!)



You see why science has all these strict rules? And why people should follow them? If you don't, huge messes like this one can happen, that would take years to resolve, and are sometimes even impossible to resolve. With the result: we'll never know who was doping during the 1999 Tour.




"In response to 2., EPO was being continued to be used by the peloton when tests were introduced. But they discovered methods (microdosing, overhydration) to avoid detection. The methods would not have avoided the platinum presence in urine making naturally occurring EPO "different" and being analysed as a positive."


Again, if you don't know the time detection limit for platinated EPO in urine, and you don't know what effect Pt will have on the EPO test results, you can't claim to know anything. (Did you *prove* that it gives false positives, every single time? Or, how often? Or, that it causes ambiguous results, and how often? Just what does the Pt do?)

Some percentage of your natural EPO will be platinated, some will not be. If it's long after your last dose of Pt (yes, but *how* long? We need to know!), the percentage of platinated EPO may be very small, and so the "positive" result may be dismissed as not real. Or, the presence of platinated EPO may change the results of the test in some way that we won't know about until someone does all that work. Or, Pt may make the test look ambiguous, or who knows what.


That's the whole, entire, point: we can all speculate about this forever, and all of our different opinions will have some validity! But, we will never *know*, thanks to those careless (or worse) scientists who did this research.


My scientific judgment is that I do not accept what's so far been reported about the EPO testing of the 1999 Tour as proving anything, for any of the cyclists. As scientific proofs, those results do not even exist.



Sorry for this book-length post! I really am not trying to clear Lance Armstrong, he and all the other cyclists may have been doping extensively for years. I wouldn't know about that, and, as I say, I don't care.


But, I think that it's the responsibility of scientists to emerge from their ivory towers every once in a while, and to give some info about science when something scientific is making an impact on real life, so that people can come to more informed judgments or decisions. Sorry, I don't mean that to sound pompous.


I guess I'm so ****** because all the rest of us scientists have to live up to all the picky little standards and step carefully on all the many bases (and, we spend a heck of a lot of boring time doing this). These guys can't whine that they should be exempt from the rules that everybody else has to play by. They need to put up or shut up.


Again, JMHO. Would love to know what you think!

 

[garnetstar]

Oh yeah: doesn't overhydration work by diluting your urine so much that the concentrations of all your EPO (natural, artificial, and platinated) drop to undetectable levels?

So, if you really hydrate enough, your urine should show up as having no EPO present at all? (i. e., such low levels that none can be detected?)



That would be really funny if true, because in the Nature paper, the testers report that, of their 102 samples from the 1998 Tour, 77 had undetectable levels of EPO. No EPO at all, natural or artificial.


So, I guess that more than 3/4 of the Tour samples just happened to have naturally undetectable urine EPO levels: not only no artificial EPO present, but no natural EPO present either.

Statistically very unlikely, and, really, very suspicious!



So, if you can totally get rid of *all* the detectable EPO in your urine just by, say, hydrating a lot........then why bother to find a test that'll distinguish between natural and artificial EPO? No cyclist with a head on his shoulders is ever going to turn in a urine sample with any detectable EPO in it at all, so you won't ever have any opportunity to discover whether that non-existent EPO is natural, platinated, or artificial!

Why turn in a urine sample that has any EPO in it at all, even if you're clean? The test might give an ambiguous result, or even a false positive! (And, it might get leaked to some newspaper.) Why take a chance, just don't give them anything to test! It's not illegal to have *no* detectable EPO in your urine, I would think.


I am so shrieking with laughter about this, so I really hope that my understanding of how overhydrating works is correct, and my mirth justified.....


(The "microdosing" you mention, if it means what the name sounds like, wouldn't rid your urine of all your natural EPO, right? It would just keep the levels of artificial EPO in your urine very low. I guess that those riders who provided the 1998 Tour samples must have been drinking their heads off, to so completely vanish *all* traces of even the natural EPO that should have been there, let alone the artificial EPO that probably was!!)


Actually, I think that there's another way that the cyclists could have been completely vanishing their natural EPO, and they'd have to use artificial EPO to do it! Also, it seems like the testers, not the cyclists, were the first ones to discover this method, and publicize it! But, later....

 

[VeloFlash]

1998 and 1999

No tests for EPO (introduced TdF 2001). Riders only used EPO test avoidance techniques from 2001.

Riders were only concerned with "health" test where they were rested if hct level exceeded 50% (unless exemption produced).

French urine test only detected the use of EPO within 24-36 hours of use. This was the test used retrospectively on 1998-99 samples in 2004. Those 77 "clean" tests were either through non EPO use or EPO could not be detected as its administration was outside the 36 hour window.

 

[garnetstar]

1998 and 1999

No tests for EPO (introduced TdF 2001). Riders only used EPO test avoidance techniques from 2001.

Riders were only concerned with "health" test where they were rested if hct level exceeded 50% (unless exemption produced).

French urine test only detected the use of EPO within 24-36 hours of use. This was the test used retrospectively on 1998-99 samples in 2004. Those 77 "clean" tests were either through non EPO use or EPO could not be detected as its administration was outside the 36 hour window.

VeloFlash,

Thanks for the info! (And, BTW, you make arguments *very* logically! If you should ever want to leave whatever your current job is, you have a bright future as a scientist ahead of you.)


As I say, it looks like there's a way that, if the 1998 Tour cyclists were doping, they would have undetectable levels of *natural* EPO in their urine, and it looks like they wouldn't have even known about it! I'll post something about that sometime. So, it wouldn't have mattered if the cyclists were using test avoidance methods in 1998-2000 or not, they would have been unknowingly vanishing their natural EPO levels just by doping (very ironic!).

Of course, as you say, there'd be no *artificial* EPO for the 2000 tests to find, if the last dose of it had happened to have been administered before the time-detection limit 24 -36 hour window. (Or, if the cyclists weren't doping, which most of you people who know about cycling seem to think is unlikely.)


But, the cyclists wouldn't have known in 1998 that there was a test coming that had a one-to-three-day window, would they? So there's no reason that they would have avoided injecting EPO anytime they felt like it. (Do you think that's right reasoning?) I'll explain later why that is significant, if it's true.


The tests on the 1998 Tour samples that were scientifically published were done in 2000. (Or else, before 2000. I don't know when they started their research.) That's the date on the Nature paper in which they report their 1998 Tour results: June 8th, 2000.

(They may have done some more tests on more 1998 Tour samples in 2004, for all I know, but these results don't seem to have been published. I can't find them, anyway. And, as I say, unless they've been published in a journal, scientifically, they don't exist.)


No one has yet published scientifically when the tests on the 1999 Tour samples were done! The lab claims that it was in 2004, I suppose we ought to give them the benefit of the doubt on that (it's probably true), but I want to see it in the scientific literature before I'll be certain. These guys don't exactly have a perfect track record for excellent carefulness.



But, what I find so suspicious about the 77 out of 102 1998 samples is that there was no detectable *natural* EPO in the cyclist's urine. (*Not* artificial. As you say, there are valid reasons why no artificial EPO may not have been found in every sample.)


A statistically significant number of people, the majority in fact, should have detectable levels of natural EPO in their urine. That's why this test has to try so hard to separate the natural from the artificial EPO signals--almost everyone's urine sample shows the presence of natural EPO. Your body is producing it all the time, right? So, it should be there, and should almost always be detectable.


So, why did more than 3/4 of the 1998 Tour samples show *no* detectable natural EPO? The researchers say that any EPO present in 1998 would still be there in 2000 (if the samples were well-preserved), and I believe them, for now, anyway. That, from the results in their Nature paper, I think should have a greater than 50% likelihood of being true, than not.


Of course, no one's published any data on whether EPO levels decay (or whether artificial slowly changes to natural or vice versa, or whether natural decays faster than artificial, or anything about what might happen to EPO) over time! So, don't take my current opinion that the 1998 samples were still the same in 2000 as they were in 1998, as gospel truth. It's just what I think is most likely the case, with the info that I have right now.

I do agree with the people who are protesting the validity of the 2000 or later retroactive results: before you claim that retroactive testing is totally reliable, you've got to *prove* what the lifetime and characteristics of preserved EPO are. You can't just say, well, that's what we, the experienced researchers, think! Play by the rules, guys!


Anyway, if, as you say, the cyclists didn't know, or didn't try, to overhydrate from 1998 - 2000, I think that the best scientific explanation for the way-too-many samples' lack of natural EPO is: that some of the cyclists *were* in fact, doping! The testers *own results* (published in another paper in 2002, I'll get this one posted as soon as I can, so you can decide if what I'm claiming seems reasonable to you) seem to prove this!


The testers may not have noticed that their 2002 data seems to indicate some cyclists' doping in 1998 (there's a *lot* of things that my own data have indicated that I haven't noticed, till someone else pointed it out to me!).

Or, the testers may not have wanted to publish this idea, since it's only inferential, not proved by specifically-designed conclusive experiments yet. Or, if the testers did try to publish this theory, maybe the paper's reviewers tossed it out, for the same reason.


The testers' 2002 data is *consistent* with the theory that some 1998 tour cyclists were doping, but not yet *conclusive* of that theory. It's JMHO that their results suggest some doping as the best current scientific explanation for the mysterious lack of natural EPO in the 1998 Tour samples.


And, of course, if I want you guys to really believe me, I'll have to publish conclusive experiments in the scientific literature, won't I? If I'm complaining so much that the testers aren't playing by the rules, then I sure have to play right too!


Later! (I have to go to work now. I do sometimes think about my own science, instead of just criticizing other people’s.)