Sunday, October 28, 2007

A Dangling Issue

In the comment thread on M's post, M re-raises a dangling issue that has not been discussed recently:

OMJ concede that was a possibility but he would want to know what special substance was in Landis' urine that it would always coelute at the same RT. Remember there were positives for multiple stages. He thought it was very unlikely such a substance existed, and it was up to Landis to come up with some plausible substance. We know from the blank urine sample that there was no normal urine substance that eluted at those RTs.

We can see by inspection of the chromatographs that Landis' samples were "dirtier" than the blanks, so we know there is more stuff present.

Julich has often raised the point that Landis should propose something that might be a consistent co-elute, and this becomes a legal question. Who has such a burden? Does Landis need to prove something else was there, or should the ADA prove something else isn't there?

We think there is a tragic hole in the Rules, where there is no discussion at all of the purity of the peaks. One reading is that it purity is assumed by good practice, but as we know, if it isn't written down, it is tricky to get a Panel to accept what appears to be common sense.

As a practical matter, it is impossible for Landis to run the kind of research tests that might make the demonstration Julich requests. The circumstances that lead to co-elutes are highly specific to the particular laboratory setup and chemistry. LNDD has not provided the details of the chemistry, and they will not run experiments for Landis in any event, per the "Ethics" code. And should they do so, who's to say if they'd do a repeatable job of it.

As a theoretical matter, it has seemed to us for a long time there is a very plausible candidate: metabolites of the cortisone Landis had injected for his hip.

We note that the LNDD never really seemed to have identified the presence of this substance that required a TUE; nor did it make the slightest effort to rule out any effect it might have had, even though it was clearly identified as present in the doping control form.

As historical note, Duckstrap looked at this in March:
At DPF, Duckstrap sees two dots that might connect -- cortisol metabolites and unidentified peaks that may have skewed the carbon isotope ratio.

A different discussion is looping around the issue of whether Landis' cortisone was detected, or not detected, reported or not reported or what. What seems agreed is that the data shows it was detected in at least some tests, but it seems not to have been explicitly noted for reasons unclear -- perhaps there was an undocumented discretionary choice not to follow up, or there was an undocumented TUE evaluation. Rare participant Dumas is asking some interesting questions and bringing in good new information.

And slightly later he thought the numbers were suspiciously similar:
Just to add further fuel to the speculative fire, I will point out that metabolites of cortisol have four acetylation sites, instead of the two found on metabolites of T. The number of carbons in cortisol metabolites is 21 vs. the 19 in T metabolites. If a peak containing primarily cortisol metabolites were a significant contaminant or were misidentified, then because of the highly negative CIR of the acetic anhydride used for derivitization (-53 per p. 351), a cortisol metabolite with a corrected CIR of -24 would have an uncorrected value of -32 in the IRMS, exactly in line with the observed CIR of the mystery "5bA" in the Landis F2 sample, and with the putative "5aA" sample in his F3 sample.

This has always been TBV's favorite theory, and is what we wrote in Our Appeal Brief as a scenario that could be offered:
  1. The chemistry is bad
  2. They don't know what they are looking at.
  3. There are likely contaminants skewing the results.

28 comments:

Phosphorylasekinase said...

I like this cortisone theory, but isn´t the chemical preparation eliminating all substances other than T-metabolites? (Again in theory, because we see a lot of other stuff in the fractions).
Is there anybody who knows if there is a possibility cortisone metabolites could have shon up in the GC/MS and GC/C/IRMS runs despite the sample preparation?
Thanks.

DBrower said...

That is kind of the question, isn't it?

The "bad chromatography" oft-mentioned means: inadequate chemical separation leading to noise, excess peaks, and co-elutes.

LNDD doesn't appear to have done anything to rule out co-elutes in the IRMS at all, because there's no mass-spectra available; and we know they didn't look for anything in the GCMS except what they were looking for.

This is why it's an appealing theory: It ties together a bunch of open issues in a way that is consistent with all the reported results. Should there be a burden flip, I don't know how the ADA could refute it. That is why the establishment of a burden flip is important.

TBV

Phosphorylasekinase said...

I agree that a flip of burden helps, but before arguing that it could have been the cortisone one has to be sure that that is a realistic possibility. We do not know what the substances in the chromatograms are, maybe they are preparation substances or simply dirt or indeed cortisone. One has to know more about this...

Larry said...

tbv -

I'm not sure why you see this as a legal issue. Sounds like a science issue to me.

When it comes to the science, I'm not afraid of asking the dumb questions. So I keep asking, why are there more peaks on the GC/MS and GC/IRMS graphs for the FL samples than there are for the reference samples? The answer that keeps coming back is that urine is a "dirty matrix" and that I need not be concerned with all those other peaks. I've never been satisfied with those answers. Yes, of course urine is a dirty matrix; I'm well aware of the fact that there's plenty of stuff in human urine. But we're using chemistry to prepare the urine samples for the GC, plus the GC is supposed to separate out compounds. So ... what IS all that other stuff?

Again, not being a science guy, I'd say that the other peaks represent other stuff, but I can't begin to know what that other stuff might be. Are they OTHER metabolites of testosterone? Are they substances unrelated to testosterone that got caught up in the GC pulse we want to examine (in other words, the GC pulse does not produce a completely pure pulse of testosterone molecules; other molecules can get swept up in the pulse as well)? I've asked this question too, and no one knows, and I have the impresssion that the question is of no great interest to anyone but me!

Another great "black box" is the chemistry that's performed on a urine sample before it's run through the GC. I understand very little about this chemistry, and this chemistry did not appear to be a focus at the arbitration. But I gather that this chemistry is necessary because urine is a "dirty matrix", and you use the chemistry to clean up the matrix to a certain extent.

The chemistry performed by LNDD is described in their SOP, and even if we had the SOP in front of us, we'd need a French chemist to interpret it for us. Based on the arbitration transcript, it took LNDD about a day and a half to complete this chemistry. I don't know if the chemistry is lengthy because it's a complicated process involving many steps, or because it's a simple process involving a few steps that take a long time to run.

One basic question is: does the LNDD SOP for its chemistry contain any steps designed to remove cortisone from the urine sample? As far as I know, this question never came up at trial. Obviously, this was a question that the FL team thought about asking at a certain point -- it's a pretty obvious question. They must have thought it would not help FL's case to raise this question, but we have no way to know why they thought this.

Another basic question: why didn't any of the science experts available to FL run tests on a spiked sample containing testosterone and cortisone, to see what would happen? If we're using a spiked sample, we're putting the chemistry issues to one side, so we don't need to know what chemistry was used by LNDD. All you would need, presumably, is access to the same test equipment used by LNDD. It doesn't sound like an impossible test to perform. However, even OMJ thinks it would be expensive for FL to do the kind of testing that would prove (or disprove) that cortisone might interfere with testosterone in GC/IRMS testing. So again, I seem to be running into a dead end.

In any event ... I don't see this as a legal question. We don't have any evidence (other than hypotetical conjecture) that cortisone might interfere with testosterone in these tests. So, what steps could we possible require on the legal side to prevent this interference?

The best thing we might come up with are hard standards for what constitutes "good" and "bad" chromatography. You've stated elsewhere that there's a connection between bad chemistry and bad chromatography. IF the FL test results were produced by cortisone-testosterone interference, and IF this interference was the result of bad chemistry, then perhaps the bad chemistry would reveal itself in bad chromatography. The reasoning here is not exactly direct, but I can think of nothing better at the moment.

Can we come up with good legal standards for what constitutes good chromatography? The standards I heard discussed during the trial were fuzzy at best. So you had the FL witnesses condemn the chromatography and the USADA witnesses give the chromatography a passing grade, and there was no hard standard that either side could use to judge the conclusion reached by the other side. But, perhaps it's possible in theory to come up with objective standards for chromatography, similar to the 2007 standards developed by WADA for reading EPO tests.

Do you have any other ideas that we lawyers could use to develop the kinds of rules you'd like to see here?

Unknown said...

An interesting discussion. Thank you M, Larry, TbV, and everyone else that has contributed.

You, collectively, already seem to have done a more logical and reasoned analysis of the Malibu hearing than the arbs.

There are a number of intelligent folks working on it here. Campbell might be excused for a few misses due to working largely alone. Not sure what the Majority Award duo, plus one, were up to? They had several heads working on this over an extended period of time. M's analysis, while it sides with the Majority, seems to be better reasoned IMO.

In any case, thanks to all and keep at it!

DBrower said...

Larry,

what seems to me the (first) legal question is whether it is Landis' burden to prove an alternate theory, or the ADA's to prove such a theory could not have occurred. This isn't science, but a somewhat critical procedural issue in the case as we stand.

The second, deeper legal question is the generalization of the first. The IDCR says what you are supposed to do to identify some of the substance in the peak. It does nothing to say there should be nothing else in the peak, and what suffices to demonstrate the purity of the peak. This is a omission in the standards, and there should be something. In its absence, we seem free to conclude there is no requirement of the sort, which is the position I'd expect to see argued in the case by the ADA; or that the standard implies purity or the science is meaningless, and the absence of the proof is a violation, which I'd expect Landis to argue. Logically, the Landis position seems correct, but a literal reading of the rule would go the other way. It's an example of where "strict constructionist" reading leads to a conclusion that is "correct" because of flaw in the rules, not by the science.

I think your inference is essentially right, that bad chemistry reveals itself in bad chromotography: overlapped peaks, dirty baselines, unclear separation, shoulders. This is why multiple measurements are taken, and should be taken, to make sure what you think you are seeing with one measure isn't something else entirely.

In the instant case, the data that seems not to have been well-collected, evaluated, or presented for independant judgement is the mass-spec data of the IRMS, which would indicate whether what was in the peaks whose CIR was being measured is what is only what is believed to be there.

TBV

DBrower said...

I see that 5.4.4.2.1 says, "The method should avoid
interference in the detection of Prohibited Substances or their Metabolites or Markers by components of the sample matrix.
"

Matrix interference, I think, is another way of saying "co-elutes", so this section does seem to say they should know there is nothing else present in the peak. Unfortunately, it offers no concrete criteria for making that determination. I suppose that means that Landis can say absence of any proof of no matrix interference is an ISL violation on this point, but I don't know how far that can fly with out some more specific grounds. It does mean it's not legally hopeless to argue the co-elute theory, since there is ISL that says there should not be matrix interference.

TBV

m said...

TBV,

I see this as the little green men theory. If there is no evidence, one can never exclude that the little green men put something in the sample.

Landis has the burden here to present some evidence.

So at a minimum, Landis should show that the cortisone metabolites elute at the same time as one of the testosterone metabolites. This could easily be shown if true. In fact one can probably look it up in a reference library. Similarly Landis could have run a test on his own urine. Why didn't he do that? Because he was afraid of what he might find out, and he would have to reveal it in discovery?

Finally, OMJ says that the cortisone metabolites even if present would lower the carbon ratio delta, not increase it. Further, RH says that cortisone doesn't affect testosterone metabolism, so would not have affected the testosterone results.

So I think this is a loser, unless Landis presents new evidence, which he easily could do. He's spent $2mil and a new IRMS test is what $2000?

blackmingo said...

TBV,

I know it is reported that Floyd got intra-articular cortisone injections, but do we know for sure that it was cortisol/hydrocortisone or another steroid? Has Floyd confirmed with the health care provider giving the injection that it really was hydrocortisone?

My suspicion is that he did not actually have cortisol/hydrocortisone intra-articularly, but one of the glucocorticoids that are more commonly used (triamcinolone or methylprednisolone) for this purpose.

Any hope of clarifying with Floyd or the health care provider that performed the intra-articular injection what exactly what was injected into his hip? I am assuming we could then take this compound, find out how it is metabolized and in what form is it found in the urine, and then guess whether it might coelute with 5aA.

Dan

DBrower said...

When something is attacked with perjoratives like a "little green men theory", I take that as a sign there may be something there that is being reflexively dismissed in hopes it will go away.

There is evidence in that LNDD hasn't proven there wasn't matrix interference. ISL 5.4.4.2.1.

As noted in the post, co-elution is very sensitive to chemistry and instrument setup, and Landis has never been in a position to have LNDD (or any other WADA lab) run experiments on his behalf. This is WADA "ethics" in action. No other lab does things the same way as LNDD, so doing it elsewhere wouldn't prove anything.

If there were library data available, it probably would have been dug out by someone by now.

The claim that cortisone would lower the delta makes no sense, as the cortisone is much more highly negative; see Duckstraps comments above. I don't take anything RH has to say as definitive, and in this case it is completely irrelevant, because we aren't talking about cortisone afftecting the testosterone metabolisation. We're talking about T metabolites co-eluting with C metabolites, which is a completely different thing. RH is wrong, and intentionally obfuscating the point.

I'll believe it's a losing argument when someone who is qualified and hasn't already taken a side on the issue explains why it is unlikely.

I don't believe the details of Landis' medication has been revealed; the days of the injections are known, but I don't have them handy.

TBV

DBrower said...

Mingo, getting the information you want would require Floyd to allow it (medical information), and to find the administering physician for the details. I'll ask, but don't expect a quick response.

TBV

m said...

TBV,

"As noted in the post, co-elution is very sensitive to chemistry and instrument setup, "

That may be or not. But Landis could show under some standard conditions that cortisone and testosterone metabolites elute at the same time. Those values would be in the reference library. That would at least present the possibility that cortisone would co-elute. So why didn't he present that?

The little green men was to show that you set up an impossible burden for anyone to prove. It's up to Landis to present some evidence that the little green men (cortisone) is present in this case.

m said...

TBV,

Plus, if this is such a strong theory, did Landis push it in his final summation?, and why didn't any of the arbitrators discuss this at all.

Larry said...

TBV -

I've already proved beyond a shadow of a doubt that FL is innocent based on TD2003IDCR. NOW you want me to prove him innocent based on LNDD's failure to prove that it avoided interference under ISL 5.4.4.2.1? How many times in one weekend do I have to prove that FL is innocent?

OK. Just kidding.

For your first legal question, you ask if the ADA should be required to prove that no interference could have occurred. Practically speaking, how could they do that? It's very hard to prove that something did not happen.

For your second, deeper legal question, you say that the ISL should require purity in the peaks used to convict an athlete of doping. That's a great idea, but how do we prove that a peak is pure? We can't measure peak purity directly. All we can do is examine the chemistry used to prepare the sample and the resulting chromatography. (Or am I missing something?)

I have no idea what constitutes good chemistry, since no one ever seems to talk about it. But we have lots of discussion on what constitutes good chromatography. If peaks are well-formed and well-separated, we have more confidence in the peaks. If they look like the peaks for the FL samples ... then we have less confidence. The athlete will hire his experts, the ADAs their experts, and the experts will disagree, since we have no agreed-upon criteria for what constitutes good chromatography.

There might be more science to this, but I doubt it. We can't measure peak purity directly. We don't seem to have any criteria for the chemistry or the chromatography. Under these circumstances, if we placed the burden on the ADAs to prove peak purity, we'd never have a single doping conviction. If we place the burden on the athlete, then the athlete will also lose ... unless he can come up with another winning theory (such as, for example, that the lab failed to adopt criteria to identify IRMS peaks).

I don't know what else I can say. All we're discussing here seems to go hand-in-hand with GC testing. I've wondered privately whether we ought to rely on GC testing as a means of determining whether an athlete like FL should be condemned for life as a cheater. I think that THIS question is a legal question: is GC testing so complicated and so uncertain that the use of GC testing is somehow a violation of an athlete's rights - to the athlete's right to due process, perhaps.

But so long as we're determined to use GC testing as our primary anti-doping tool, then IMHO the questions you've raised about peak purity are not legal questions. They are science questions. When the science answers the questions, and provides us with criteria that can be used to measure peak purity ... then the application of these criteria to a given situation will be a legal question.

DBrower said...

Now we're at the legal question: Must Landis show a theory, and prove it, or not? It appears at hearing they chose not to, believing that an ISL violation on identification was sufficient.

Assuming for sake of argument that Landis can show there was no proof of the absence of matrix interference, does he need to prove anything else?

If the mass-spec data that would prove the presence of another substance was available, the question might be resolved, but it has been destroyed for the S17 sample, and not made available for the other B samples. Perhaps it has been destroyed for them as well. How convenient.

Where does the WADA code say Landis must pull a Perry Mason and produce the real killer?

I admit it would be nice for the movie to get that sense of closure.
I am also in favor of presenting some plausible theories in the argument, but I don't think they need to be "proven". I think it is the ADA's job to prove what they have done is correct, and if they can't disprove the theories that coincide with an ISL violation, they should lose.

There is nothing to prevent USADA from running additional tests to try to prove that cortisone metabolites don't co-elute using LNDD's chemistry and instruments. Why don't they do that? It's easy for them, and they'd shut this particular line of argument down.

But even so, they won't have gotten rid of the unknown co-elute theory unless they produce mass-spec data for S17 sample (which they can't), or try to discredit it with the mass-spec from the alternate B samples that were "positive" on the "5aA."

TBV

DBrower said...

Larry,

The data that would resolve the peak-purity issue is the mass-spec data collected at the time of the IRMS. This is not the GCMS data; it is equivalent to the "three ion" monitoring of the T/E test, which LNDD also did not do properly.

The S17 mass spec data was on the drive that was wiped in February, and LNDD conveniently did not back it up on the CD they gave Botre.

There is no clue whether it was collected, evaluated, or remains in existence for the alternate B samples.

It would be really good if the ISL required it to be collected and evaluated the same way the three ion rule is in place for the T/E test. But it isn't, we have only a mealy-mouthed statement that matrix interference should be avoided.

TBV

Mike Solberg said...

TbV 10:02, talk about getting snippy...

syi

Mike Solberg said...

(I read quickly but I don't think anybody said this yet...),

m, while normally impossible, this is one case were you can easily prove a negative. It is easy to prove that there was no cortisol, or anything else in those IRMS samples: it's called the complete mass spec data, and LNDD either didn't collect, lost, or buried that information.

To quote your favorite DPF poster on March 30 (obviously pre-hearing)...OMJ wrote:

"What is possibly relevant is that there is a sht-load of material in this F2 peak, so it is reasonable to speculate that some of it might not have been extracted into F2, but might have spilled over, so to speak, into F3, where it could interfere with the 5bA peak. I agree that could happen (and in fact, if there is no 5bA in that F2 peak it means that more of that peak is contaminant, and there is a greater likelihood of spillover), I just don't see any evidence so far that it has. As always, the complete mass spectra would settle that issue—and if it turns out the lab does not in fact have those spectra, I will scream foul as loudly as anyone else here."

I don't recall ever hearing the scream.

syi

Mike Solberg said...

Okay, so TbV did say it...duh.

syi

Larry said...

TBV -

Between your recent posts and Mike S.'s posts, perhaps we might try to make the legal argument that (1) ISL 5.4.4.2.1 requires the WADA labs to preserve the mass-spec data collected at the time of the IRMS, (2) LNDD's failure to do so is an ISL departure, (3) USADA thus had the burden of proving that this departure did not cause the adverse finding, and (4) USADA failed to meet this burden.

For the moment, at least, I'm NOT saying that this is a fair reading of the ISL. For the moment, I'm reserving judgment. However, this reading DOES at least put the question we're discussing into a more-or-less specific legal context (more specific, in any event, than the due process argument I floated yesterday).

However, TBV, notice that in my legal argument, I never used the word "cortisone". The only way that cortisone enters into this legal discussion is, possibly, if the parties argue point (3): the ADA might say something like, what interferences could there be, and the athlete would respond, any number of interfereces are possible, including cortisone.

I did not like the analogy to "little green men". But I share M's discomfort with requiring the ADAs to prove that an adverse doping finding could not have been caused by the presence in the athlete's system of a specific permitted substance. As a general matter, we KNOW that this is the case. We KNOW that interference is a theoretical possibility. But we've chosen to rely on mass spectrometry as a valid means of testing. If mass spectrometry IS valid, it's unreasonable to ask the ADAs to consider every substance that might reasonably be suspected to be in an athlete's body and prove that the substance could not have caused interference ... or that the lab took specific steps to prevent such interference. That is, unless you can show me science that would make it reasonably possible to supply such proof.

Consider breathalyzer tests. This is a technology that is considered to be reasonably reliable, but it can screw up. Good DUI attorneys (and I'm NOT a DUI attorney) know all about this, and they'll ask the kinds of questions that might lead a court to throw out a DUI case: for example, did the defendant have a reflux condition? That's a condition that's known to throw off a breathalyzer reading.

But even in a DUI case, where the burden of proof on the prosecution is higher than that imposed on the ADAs, the prosecution is NOT required to walk through a checklist of items that might affect a breathalyzer test and prove that none of them were present. It's up to the defense attorney to raise these questions.

FL had good counsel. Also, FL is in a better position as a general matter to know what substances were present in his system. I think it's fair to say that if FL never raised the issue of cortisone interference at the arbitration, then USADA should not have been required to prove that the adverse finding was not caused by cortisone interference.

The issue of the IRMS mass-spec data is a different and more general issue that we could consider separately. Let me ask a science question: if we HAD this data, what would it look like? Would we be back to making the same kinds of "fuzzy" arguments on whether this data was "good" spectroscopy or "bad" spectroscopy?

DBrower said...

OK, let's say that naming cortisone specifically is a red herring (that is interesting to some). It is clear that the Landis team did raise the issue of possible co-elutes without naming them.

(I suggested earlier they may have felt no need to identify potential co-elutes, on the belief that they had no burden to do so).

Rather than the intoxication analogy, I'd say a more appropriate traffice enforcement metaphor is a speeding ticket. The requirement that there be no matrix interference is like a radar rule that there should not be conflicting target vehicles. Say there is someone driving a 1966 VW microbus uphill in traffic and being passe. If he's given a ticket for going 90, he can reasoanbly suspect Officer Friendly's gun was pointed at some other vehicle. Does he need to prove which other vehicle it was, or is it enough to indicate the likelihood that a mistake was made?

The mass spec data would reveal truth just like the three-ion monitoring in the T/E test; Similarly, video tape from the cop's car at the same time as the radar shoot would tell whether there was conflicting traffic.

TBV

Mike Solberg said...

TbV, all your comments about 5.4.4.2.1 recall, of course, the majority's discussion of this in their final decision. I thought when I first read it, and I think even more strongly now, that it was a terrible part of their decision. It was almost farcical.

Sorry, this will get long, but I think it's worth remembering it accurately:

About 5.4.4.2.1 the majority decision says (paragraph 240):

"While the Panel acknowledges that it may be ideal to have “perfect
chromatography” where there is absolutely no matrix interference and no coeluting peaks, or sloping baselines it is also mindful of the fact that this is not always possible, particularly when dealing with human samples obtained in less
than ideal circumstances. The LNDD’s chromatography was according to the experts called by the Claimant good to very good. Although it perhaps could have been better it remains “fit for the purpose” AND UNQUESTIONABLY INDICATES THE PRESENCE OF EXOGENOUS TESTOSTERONE IN THE RESPONDENT'S "A" AND "B" SAMPLES (see (1) below) In applying the language of the ISL what is required is that the “method should avoid interference.” THE LANGUAGE IS NOT MANDATORY. (2) Had the drafters intended that matrix interference be avoided it would require wording such as “shall” or “must”. For this Panel to accept the submissions of the Respondent that matrix interference must be avoided would be a misconstruction of ISL 5.4.4.2.1. DR. AYOTTE CONFIRMS THIS STATEMENT IN NOTING THAT A LABORATORY DOES NOT VIOLATE ARTICLE 5.4.4.2.1 OF THE ISL JUST BECAUSE IT PRODUCES A CHROMATOGRAM THAT CONTAINS MATRIX INTERFERENCE. (3) Therefore, EVEN WHERE MATRIX INTERFERENCE HAS OCCURRED IN THE STAGE 17 CHROMATOGRAMS IT WOULD NOT AMOUNT TO A VIOLATION OF THE ISL (4).

(1) Doesn't this indicate a basic misunderstanding of the difference between the GCMS and the GC/C/IRMS? The IRMS doesn't and can't "indicate the presence of exogenous testosterone" - all it knows is that there is SOMETHING in here with a CIR of such-and-such. You have to connect it, somehow, to the GCMS to determine what it is.

(2) I find this discussion of the "should" language ridiculous. Yes, they found a "should" in this section of the ISL. There are lots of "shall"s and "must"s also.

5.4.4.1 says:

"5.4.4.1 Selection of Methods -
Standard methods are generally not available for Doping Control
analyses. The Laboratory SHALL develop, validate, and document in-house methods for compounds present on the Prohibited List and for related substances. The methods SHALL be selected and validated so they are fit for the purpose."

Now, for this to make sense in the context of LNDD's methods, it has to apply to the whole process of GCMS and GC/C/IRMS together. You need BOTH to show you have ENDOGENOUS testosterone. You can't just say that they have met the standard with the GCMS alone, because all they have done with that is identified that testosterone is present, not whether is exogenous or endogenous. If they haven't connected the GCMS to the IRMS, they haven't met the standard, and it's "shall" language.

5.4.4.2.1 says:

"Confirmation methods for Non-threshold Substances MUST be validated. Examples of factors relevant to determining if the method is fit for the purpose are..."

At which point the majority decision skips to the "matrix interference" bullet point and its "should" language.

But what about the "specificity" bullet point and its "must" language? To wit:

"Specificity. The ability of the assay to detect only the substance of interest MUST be determined and documented. The assay MUST be able to discriminate between compounds of closely related structures."

Again, this has to apply to the whole process, the combination of the GCMS and the GC/C/IRMS, because with just the GCMS you have not identified a PROHIBITED substance, just testosterone from any source.

(3) While technically true, look at her actual testimony (transcript p. 852, pdf p. 689):

"A. I said that the section was dealing with validation of method, and that it was important to be fit for purpose that the method employed and developed by the lab would be checked for specificity, identification, capability, robustness, carryover, matrix interference and standard. My testimony was that if there is,
at one point, a matrix interference in the course of the testing, that does not make it a violation of the standard.

Q. IT JUST MIGHT MAKE IT INACCURATE, RIGHT?

A. If you are -- YES, EXACTLY. YES.

Q. Just so that I can be clear, when you're looking at a laboratory document package for what was actually done on the case,
certainly you want to check specificity, correct?

A. Well, we have to -- no. When we
were looking at the documentation package, we don't get to verify the validation done at the lab. We want to make sure that the lab has
given a result which is -- which is specific, which is accurate.

Q. Okay. But setting aside the
validation process for a moment, and talking about the actual analysis, the -- in the
analysis, the lab -- just going through these bullet points, the lab still has to show that they properly identified the substance, correct?

A. Correct. That is why there's the
purposes of having controls in the assays as well.

Q. And during the analysis, the
laboratory still has to demonstrate that there was no carryover, correct?

A. Yes.

Q. And during the analysis, the
laboratory still has to demonstrate that there was no matrix interference, correct?

A. Correct.

Q. And if there is matrix interference in connection with the analysis of the sample, that affects the accuracy of the results, correct?

A. Of course.

Q. And of course, if there's carryover in the analysis of the sample, that affects the accuracy as well, correct?

A. Yes.

Q. So these points here don't just go to validation; these are also things that you need to look at in the analysis.

A. Yes."

WTF? It "might make it inaccurate," but not a violation of the standard? That sure sounds to me like the testimony of someone trying to be technically truthful, but trying not to be helpful, as would be expected under the "good neighbor" policy.

For the majority to pick up on her comment that it was not a violation of the ISL, but not to attend to the rest of her testimony in that same section, well, that is "not a search for the truth."

(4) Again, doesn't that reveal a basic misunderstanding of the difference between the GCMS and the GC/C/IRMS? If matrix interference has occurred in the IRMS part of the process, then there is no way to know what all was contributing to the CIR you calculated.

syi

p.s. sorry for all the CAPS, but the html tags aren't working for me.

Larry said...

TBV, actually I think that the breathalyzer analogy is better than the radar gun analogy, because the first involves issues of organic chemistry that are closer to MS issues.

But the radar gun analogy will serve. In the case of the 1966 VW Microbus, I think that the court would initially be willing to rely solely on the evidence from the radar gun. It would be up to the defendant to assume some burden of proof here, notwithstanding the presumption of innocence. (Or more accurately, that the police overcame the presumption of innocence by showing the radar gun test results.)

So the defendant might say, "I was driving a VW Microbus uphill and I can't do more than 45 MPH uphill, even with my foot to the floor and a gale force tail wind." To which the judge would probably reply, "Officer Friendly, could you have made a mistake? Could you please produce the video tape?"

But the case you've posited is too easy. It would be like FL arguing, "I have a medical condition that makes it physically impossible for me to dope with testosterone. If I do it, my face swells up to the size of a basketball, and I can't ride a bicycle for a week afterwards."

So, let's use a better speeding case as an analogy. A better case might be, the driver was driving a Ferrari. 90 MPH is physically possible for the Ferrari (just as it is physically possible for FL or most any other rider to dope). The driver knows for certain that he wasn't speeding, but cannot prove that he wasn't speeding (any more than FL or any other athlete can prove he didn't dope). The driver knows that the police had a videotape as well as the radar gun results. If the videotape still exists, it might or might not exonerate the defendant.

What would have to happen to get the police to produce the videotape? If the defendant is silent, should the videotape be part of the proof that the police is required to offer to get a speeding conviction?

Before we answer that question, let's examine a side issue. If the police has evidence that it knows will exonerate the defendant, the police is required to produce the evidence. (I think the police must also produce evidence that would be helpful to the defendant, but I don't know the exact nature of this rule without looking it up.) I'd argue the same for the lab: if they have evidence proving FL is innocent, or even tending to prove that FL might be innocent, they should be required to produce that evidence. But we don't know that they have any such evidence.

The question we're asking goes beyond this side rule. We're asking whether the police should be required in all cases to produce the videotape as well as the radar gun results ... and whether the case should be dismissed if the police cannot produce the videotape.

I think the police can fairly get a conviction based just on the radar gun results. The radar gun results, like the results of mass spectroscopy tests, are considered to be reasonably reliable without the necessity of producing a second piece of corroborating evidence. So in our speeding case and our doping case, the burden would be on the defendant to throw enough doubt on the test (or the circumstances under which the test was run) to require the prosecutor to produce corroborating evidence.

In the speeding case, this burden on the defendant does not rise to the level of a "Perry Mason" confession ("I admit it! I was the one who was doing 90 MPH, not the defendant! And I'm GLAD I did it! Hah hah hah, boo hoo hoo!"). But the defendant would at least have to create a reasonable doubt.

Arguably, FL created more than a reasonable doubt that something was wrong with the LNDD's mass spectroscopy. This is what drives Judge Hue (and others ... me sometimes) completely nuts. The WADA rules are much more mechanical than anything we're going to see in a speeding case (or a DUI case, or any other kind of criminal or civil case). In the WADA world, we're limited to looking at departures from the ISL.

But going back to our main question: no, I cannot say that USADA should have legally been required to prove that FL's test results were not caused by cortisone, given the failure of FL's team to raise this question as part of its defense. (and yes, I'm aware that the preceding sentence contains a triple negative)

Larry said...

Mike, your post is terrific, and it demands a better response than I have time to give at the moment.

The point you raise goes to a fundamental problem with the majority opinion, and what I think is a common problem most of us have when we read the ISL. Many (perhaps most) of the provisions of the ISL do not directly address what a lab is required to do. Many of these provisions require the lab to adopt standards or procedures that meet the ISL requirements, and then require the lab to adhere to its standards. So, in effect, the lab is not bound by the ISL -- it is bound by its own standards, and the sufficiency of the standards is measured against the ISL.

But for whatever reason, we don't seem to have LNDD's standards.

In my post on ISL violations under TD2003IDCR, I argue that the LNDD never adopted the criteria it was required to adopt by the TD. I think that my argument there is bulletproof and irrefutable! But I have to admit, as best as I can tell, no one ever asked LNDD to produce these criteria. Maybe the WADA rules allow these criteria to remain hidden. I don't know. All I know is that the parties spent a lot of time arguing the rules governing what these criteria are supposed to be (TD2003IDCR), and no time examining the criteria themselves (or even, questioning whether there were any criteria).

I think you're running into this same problem when you look at the ISL rules governing chomatography. I haven't looked at every rule carefully, but at least some of these rules aren't telling LNDD how to do its chromatography. They're telling LNDD that it has to adopt its own chromatography standards, and these standards must satisfy the rules set forth in the ISL.

In other words, in many cases, the ISL rules are "one step removed" from what the LNDD is required to do.

This "one step removed" business helps explain the "should" provisions of the ISL -- the provisions that the majority felt free to ignore. Imagine for the moment that you're writing rules that a city would have to use when it wrote a housing code. Rule 1 might be: the code MUST require houses to be built to keep the rain out. Rule 1a might say, the code SHOULD require that every house have a roof. Why say "should" in rule 1a? Well, maybe the writers of the city code can up with some way other than a roof to keep out the rain. Then why include rule 1a at all? Because ... having a roof is the accepted way to keep out the rain, and if your code does not require a roof, then we're going to look DAMN HARD at whatever else you come up with in your code as a means to keep out the rain.

I'll write more later, I hope. In the meantime, please consider whether the ISL rules you're trying to apply are "one step removed" in the manner I've described.

DBrower said...

Larry,

This has already wandered and I didn't mean to distract from your other post.

But...

I think there is still some mileage in the speeding analogy. You put the cop on the stand:

Q: If there were other vehicles present, might they affect your readings?

A: possibly.

Q: Were there other vehicles present?

A: There were no other vehicles present in the lane.

Q: Were there vehicles in other lanes?

A: I don't know, I wasn't looking at other lanes.

Q: If there were vehicles in other lanes, might they affect the radar reading?

A: I suppose hypothetically, but I don't have any reason to think there were vehicles in the other lanes.

Q: And you didn't look in the other lanes to be sure?

A: No. The video in the car might have recorded the other lanes, but I didn't look at it.

Q: Can we see the video?

A: I didn't think anyone needed it, so I erased it.

Q: When the ticket was issued, and Mr. X signed it, he wrote 'preserve the video' didn't he?

A: Yes, but we decided he wasn't entitled to it, and didn't need it. The radar is infallible.

--

To your later point, if most of the ISL rules are "one step removed", how could an athlete every prove a violation without access to all these other materials?

We know that the Landis discovery requests included all the SOPs, and we don't believe they were fully produced.

Anyway, this is wandering, so let's get back to the post on the violation (or not) on identification.

TBV

Larry said...

TBV -

I'm lost. Which part of our recent discussion is the part where we're wandering? (: ^ )

Yeah. The nice thing about a court is the ability to get the cop on the stand and make him squirm, to the point where the judge says "case dismissed." Not that it happens that way all that often.

On the "once removed" point, if you can't get the SOPs, all you can try to do is to argue a departure from the ISL based on a hypothetical SOP that just barely meets ISL requirements.

That sounds bad, and it's really worse than I'm making it sound, because the ISL requirements are relatively flexible (as they probably need to be). And in any case where the ISL provides flexibility, the majority arbitrators found that the LNDD was not required to do anything (based, I guess on the assumption that there was no way to determine what the lab was required to do in that case).

So if there are ways to construct alternative hypothetical minimal SOPs, the ADA will end up with the benefit of the alternative that supports the adverse finding. It's only going to be in a case where the ISL is specific (i.e., you have to measure 3 ions to prove a T/E doping violation) would you be able to prove an ISL departure.

This same thing happened with the chain of custody analysis, by the way.

DBrower said...

I think we started wandering about the time I typed, "A Dangling Issue".

Sigh.

On the other hand, I've gained new insight into the ways the ISL can't be provably violated by any lab :-(.

TBV

Larry said...

TBV -

Well, come on over to my post, where there are ISL violations everywhere you look! (at least it seems so to me)