Tuesday, May 15, 2007

Hearing - Tuesday Mongongu -- We Are Off


The new translator indicates she is stressed, needs an hour usually to review terminology and is from the Judicial Counsel of California. She will do her best. She gets encouraging applause.

She has and is doing a very nice job, given the circumstances. The parties and Panel worked together to arrange from a very reliable source, the Judicial Counsel, her attendance at the last minute and agreed to split the cost.

1st step-prep samples , 2nd GCMS analysis to ids of compounds , and the third step IRMS analysis.


Mr Dunn examines:
What are security measures at LNDD?

Need a badge to get into the first entry-use a swipe card. then go through another entry that needs a swipe card and then a swipe card into the lab. If someone from outdside the lab comes in, needs an escort.

[ This must be about tampering ]

Questions about samples:

It takes 1 day and a half to prepare the samples (this one got the previous translator in trouble) . Everyone applauds.

There are instructions and guidelines to prepare the samples.

Landis specific Sample questions:

Followed the guidelines. Followed SOP samples and instructions for "A" samples and she verified the "B" sample preparations by Ferlot. She added an internal standard in the prep. of the "A" sample. Does not include any other steroid than 5 alpha androstinal. She documented at pgs 119-122 of Exhibit 24. GCMS is the step that allows the verification of the molecules. Before any analysis, she always verifies the state of the instrument and did so for Landis "A" sample and verified tha same for Ferlots "B" sample. she knows the instrument is correct by checking the "MASS" (ion source), checks for leaks, then check standards to see if machine can check performance of the standards. She injects a solution of several molecules 5 alpha, 5 beta, keto and andro (these are acetates). The parties disagree as to terminology and the whole name has to be used so as not to confuse the compound and the target.

[ We note Brenna is pounding away at his laptop in the back of the room, with an unhappy scowl. Maybe laptops are ok? We'll check later. It would be good to be able to do the play-by-play from inside. ]

As to the Landis "A" samples; ion source checked out ok and there were no leaks and the instrument could identify the seven compounds and Ferlot's tests on the "B" samples were confirmed.

The samples were injected into the instrument for the "A" sample. It takes 3 hours to do this.
A blank urinary sample is prepared. Each fraction of the sample is tested after the corresponding fraction of the blank urinary sample is tested.

[ Botre is in the jury box, intently watching the witness. ]

A mix acetate not a Mix-cal acetate is the first injection. You can see the order of her injections, fraction of blank injected then fraction of the athlete's urine is injected.

[ Can a blank be contaminated by a preceding athlete sample? ]

3rd step-instrument verification process for the Landis "A" sample. She checks to see if the instrument is set and then stable (same quantity of gas by reference). They get specs for this. Verify precision ; that the instrument gives same measure for same analysis. Inject same solution, 3 times. 4 alkanes and it must stay within a standard deviation. She injects acetate compounds and the isotopic values were certified; they checked out-within specification. Her work was verified by Mdm Roisont. She says this is on Page 114 and 115 and he corrects 174 and 175.

Her attention is drawn to page 175, top box, and it is displayed on the screen for her. Mongongu inserts the data number 007 for the mix-cal acetate in the box. She looks to see if the number of all four compounds (?) is within accepted values (range). Then the next line is 14, and that's the mixcal at the end of the runs. She compares the differences to see if they are stable. In this case, they are good compared to the standards printed in the table below the handwritten values.

Pg 185/186 is the final analysis documentation. The document is on the screen. The isotope value, 31.64 falls outside the range on P 175. She says this is no problem because the internal standard is used to calculate the relative retention time of the molecule-it isn't exploitable because it is situated at the start of the chromatogram where there is [no? - Bill heard a "no" and TBV didn't] matrix interference. She says there is no acceptance criteria and it doesn't effect IRMS results.

[ There is some back and forth about this. Apparantly, anything will do because they don't care about it. Why measure something if you don't care? ]

Q: 31.64 is outside the range on 175. Does it affect the validity of the IRMS conclusion?
A: NO!

After quality control and the instrument checks, she injects samples into the IRMS. This takes several hours. Then she takes each data and performs a procedure to verify the background noise and data reduction. The result obtained was verfied by Mdm Roisont.

She determined the "A" sample was positive for exogenous testosterone... ando-11, ketio and Adiol. That is why the sample was reported postive. The results are in no way affected by the machine determining certain controls were out of range, even if every one of them were.

Page 186, looking at the reported results. The middle column is the reported values.
Q: What are columns 3 and 4?
A: The uncertainty ranges.
Q: Did you determine this was a positive?
A: yes.
Q: which values led to that conclusion?
A: the Andro -11K at -3.99 and the 5aA at -6.14.

Q: The table above -- what is that?
A: Values of a population that was established in the lab.

Q: do they affect the conclusion of this test?
A: No.

Q: If they were all different values, or the check marks were different, would it affect the conclusion?
A: No.

[ This is a table with some values, and oui/non columns, some of which are checked, some not. If it's to be filled out, but doesn't affect the result, what is it for? ]

Exhibit 112, LNDD 603-609
Q: What is it?
A: The operational method to verify data that is acquired.
Q: Did you follow it?
A: Yes.

Exhibit 24, USADA 124.
Q: What is it?
A: The SOP
Q: Did you follow it?
A: yes.

She verfied M. Ferlot's "B" sample results as a supervisor.

She had no idea she was doing an "A" sample analysis on Landis' sample. She isn't a fan of cycling and didn't follow the Tour (Floyd shrugs). She also didn't see Landis' medical control paperwork prior to doing the analysis, the sheet with the TUE information.

Q: Was the IRMS operating at the proper pressure?
A: yes
Q: how do you know?
A: If the pressure is wrong, the results are not stable or precise. The values would be all over the place when we analyse samples.

Q: Who made the machine?
A: It was Micromass, now GV Instruments.
Q: Did you get training from Micromass?
A: Yes.
Q: Who from?
A: Application Engineer Pascal something.
Q: Did he come to the lab?
A: Yes.
Q: Did he verify the machine was at the right pressure?
A: Yes.
Q: Same as the manual?
A: I don't understand?
Q: It came with a manual?
A: It wasn't an Isoprime manual.
Q: Anyway, the Micromass trainer indepenantly verified it was at the proper pressure?
A: Yes.

She says the pressure was correct, otherwise the results wouldn't be precise and values would be all over the place. They looked at this and determined the pressures were fine. She received training on the machine at LNDD by the machine manufacturer, a man named Pascal. Pascal determined the pressure on the ISOPrime 1, the machine the "A" and "B" sample was tested on. The machine didn't have an instruction brochure (for the IsoPrime 1). but the rep said the pressure was fine.

[ One of the Landis arguments has something to do with pressure. This has been discussed at DPF; USADA is trying to show it not meaningful, because it was blessed by a Micromass employee who visited the site. That he visited is also relevant to address any future installation problems that may be raised by Landis. ]

Sample B was analyzed by Ferlot in August, 2006. Mongongu was a verifier for the "B". She didn't break the seal (some confusion here but she says adamantly, "No"), didn't make aliquots, didn't handle aliquots, didn't process or report that data. She just verified it was correct.

[ They are trying to avoid a Landaluce problem ]

Explains positivity criteria for exogenous testosterone. She is looking for higher than 3 plus uncertainty (.8) or 3.8 difference. In this case 5 alpha, adio and 5 beta all greater than 3.8. Andro-11ketio is 3.51. She would not have declared a positive on that one. She would not have called it negative. She would have called it inconclusive. Mmsl Ferlot made that determination as well. However, the written notations include the 3.51 result. It is explained that 3.51 was superior to 3.0 and was "suspicious" and 1 was positve (6.31) and the other positive.

Q: Oh, I forgot to ask, what are the positivity criteria?
A: We declare a positive when a delta is greater than 3 plus the uncertainty, therefore 3.8
Q: If these [pointing to column 2] are greater than 3.0, you declare a positive.
A: yes.
Q: And here the 5aA are all > 3?
A: yes.
Q: the Andro -11K is -3.51 also greater than three, but -2.71 to -4.33?
A: yes.
Q: If you had only the 11k value, would it be a positive test?
A: No, I would not declare it a positive, I'd declare it inconclusive.
Q: You wouldn't report it negative?
A: No, doubtful.
Q: But you had the 5aA that let you conclude a positive?
A: exactly.
Q: And this is validated by Frelat (?)
A: Yes.
Q: Why does the last sentence say -3.51 and -6.39? Why mention the 3.51?
A: The positive is based on the 6.31, that's it. So the report is because one was positive and one was uncertain or doubtful.

[ huh? ]

They were using OS2 Optima software. She manually checked to correct error in the background noise. She manually adjusted starts and stops in peak integration because the chromatography peak.

A: "Sometimes inserted by the software in an area where there is not my peak."

[ indicating sense of ownership? ]

This was sometimes necessary to do in the Landis "A" because the results otherwise would not have been correct. On the B, M Ferlot also made manual adjustments for background and for start and stop at certain peaks. As verifying scientist it is her opinion that the adjustments enhanced data quality and reliability of the results.

Q: The procedure allows this?
A: Yes.

[ But they only use the Masslynx in automatic, so how reliable are its results, if manual fixups are needed on the Optima? Brenna said when they tried the manual button on the masslynx, the program crashed. ]

April 16 to April 27, she also participated in additional "B" sample testing from other Landis samples from the Tour de France.

Q: Were there any positive results?
A: yes.
Q: how many?
A: I don't recall exactly, perhaps four.

Q: Was the software for the April tests the same as for S17?
A: no, it was the Masslynx.

Page 84 LNDD 0725

Q: -4.96 Same positivity criteria?
A: yes.

Q: On the reanalysis [ I think reprocessing was meant ], what was the result?
A: -5.06

Break is called at 4:19 (Bill thanks God for that!)


Brian said...

He is asking her if allt he quality controls checked out when testing mr. landis' sample, and she answered: yes.

I thought that she wasn't supposed to know which rider's sample that she was testing, so how can she know that the qc checked out in Landis' specific case?

Anonymous said...

Giid question. She said (with the old translator) that she's done 400 IRMS tests. the a-sample should have been on in a long assembly line of faceless tests. So, how did she know it was his?

beeble said...
This comment has been removed by the author.
beeble said...

Obviously because that's who this whole episode is about, anonymity went out the window a long time ago.

I think you're misintrepeting the testimony if you read it as they knew who the sample belonged to at the initial testing.

Of course I've been wrong before...

Brian said...

I am not reading the testimony....i am watching it. The prosecutor asked her if she followed the quality control standard procedures while testing Floyd Landis' specific A sample. At this point, the day after the tour...his sample was to be anonymous.

Anonymous said...

And She did say she's not a fan of cycling. is she the leek trying to take down cycling? :0 Just kidding. However I would like them to ask her if she knows or suspects who is leaking the information to L'equipe. and how she knew she was working on Floyd's A.

y'all are doing an awesome job TBV, Bill, strbuk and marc.

Keep up the good work!

Atown, Tx.

Anonymous said...

oops, re read bill's play by play. I guess she didn't know it was floyds A. or did she? hmmmm.

Atown, Tx.

Anonymous said...

I'm sure she knows she followed standard procedure because she always follows standard procedure. After all, the labs are infallible and unquestionable. So she can say that with complete confidence, since it's impossible that she could have made a mistake. Wait, maybe that should be inconceivable...

OK, all sarcasm aside, I doubt they'll get much mileage out of that, unless they can somehow prove that she didn't follow procedure. Tough, when you don't have the original data files...


pem said...

Mr. Hue:

You have my praise in your coherent typing of the play by play. I know how difficult it can be trying to type AND follow what is going on at the same time and not fall behind – and especially typing with fairly good grammar. Thanks again for the updates and keeping it interesting.