Monday, May 14, 2007

Hearing Monday, Brenna Part II

Q: Masslynx reprocessing on May 4-5. About one data point, on the 5a blank that is -3.66. Is there something unique about this.

A: Yeah, that was a unique run. Moving one computer to another. If memory serves, there was a single point where Masslynx did not process things properly, or the way it did the others. At least twice, Dr. Davis said that he was happy to throw that one out.

[ Apparently, some of the blanks came up positive for doping, and this is an attempt to make that go away.]


Q: this is p 121(?) of respondents brief. Let's look at exhibit 107. Have you examined these.

A: Yes, they are correct.

Q: Have you reviewed the documentations of the samples started on April 16.

A: Yes, for 5-6 hours.

Q: What did you look for?

[ Objection: This is subject to motion to strike ]

Campbell: Does this come from LNDD?

Young: No, this was prepared from data from LNDD.

Campbell: Standard deviation?

Young: exhibit 24

[ This is part of doc pack we haven't received. ]

Have to obtain value of 3 of 4 alkanes; it's down on bottom right there. And if you want me to find uncertainty study for the 5%, I could find that too. This data comes from documents in the record.

[ Is this a gotcha that Team Landis didn't notice? ]

Q: You looked at those samples, what did you focus on.

A: The most important data was the graphical representation of the chromatograms and the baselines that the software had automatically drawn for the positives. So I looked at that and reviewed all the 5a's and 5b's and pdiols without regard to sample number first. I went through them all and looked at those three peaks...

Q: Could you tell from those doc packs whether those samples were analyzed using Masslynx or OS/2?

A: Entirely on Masslynx, so no OS/2 questions.

Q: When you looked at positives...

A: Looked at everything without regard to anything, wrote it down, as to whether the baseline was drawn reliably. Did find some weren't drawn reliably, but those don't invalidate, just mean examine more closely. Then compared positives to the list, and found that al the 5a positives were reliable. Apologize for 5b's, that may have been suspect; all the pdiols were fine.

Q: Let me show you the chromatograms respondant showed Dr. Shakleton this morning. If you'd look at exhibit 86, which is LNDD 894. This is a sample from the 18th. Is that the one you talking about having a problem with the 5b chromatogram.

A: I would have concerns for the baseline of peak 19 the 5b, but not peak 20, the 5a. The pdiol peak 22 looks fine.

Q: On the 18th, the 5b was was negative with uncertainty anyway.

A: Right.

Q: LNDD 991, which Dr. Shackleton was also asked about. What conclusiong did you draw?

A: What I see in this is peak 13 right there, which is 5b; peak 15 is 5a. The software has decided there is a peak between 13 and 15, and assigns area to what it calls peak 14. That is clearly not correct. There may be a tiny peak on the tail of 13. That would be another suspect. I said earlier the delta delta calculations are robust. My experience is that in this circumstance, one might or might not get good results from this, so I'd go back and fix it up. The software failed to identify the peak correctly.

Q: Does this affect the 5a?

A: The 5a looks to be integrated fine, just fine. We're dropping all the way to the baseline, then the value for 5a should be fine. That's what my experience tells me.

Q: That comes from July 22nd. Again, we don't care about the 5b number. Is the fact there may be not good resolution on the 5b undermine the reliability of the pdiol or the 5a data?

A: The pdiol is nowhere near it, so that's out of the picture. The only potential interaction on the 5a and 5b would be if there was big overlap, and I don't see that, so i believe the number is good.

Q: Does this affect your opinion of the S17 results?

A: No, it doesn't.

No further questions.


Q: When did you get USADA grant?

A: Believe Jan '06.

Q: USADA 1021, grant application.

A: Signed

Q How much was it?

A: One million

Q: How long was that for?

A: Three years.

Q: So you're still being paid on it?

A: Yes?

Q: Your lab is not accredited?

A: Correct.

Q: We started with quality control at LNDD.

A: Yes

Q: You said you thought LNDD QC was impressive.

A: Yes.

Q: Let me turn you attention to one standard that was added. Have you seen this before, USADA 119

A: Yes, but I don't remember.

Q: Look through 119 to 122. These shows the 5a acetate as added to sample A; turn to 121. See where it says SI 200 ng, thats where it was the internal standard was added.

A: Correct.

Q: Now USADA 299, 300 and 301 and 302, exhibit 25. That appears to be where they added it to the B sample in the 200ng.

A: It's french, I can't say.

Q: Now USADA 354, it's the internal standard, right?

A: Yes.

Q: Internal standard is added to quantify whether the instrument is correct.

A: Yes.

Q: on 354, can you explain the +5 and -5 mean?

A:: I took those to be the uncertainty limits.

Q: the instrument should find to substantiate the conclusion the instrument is accurate, fair?


Q: USADA 175, no reason to be different than 354, right?

A: Right.

Q: Please turn your attention to USADA 185. Do you recognize this page? This is the results of the measurement of the internal standard. Can you at the F1 fraction, sample A, what do you see as the value?

A: Sample A, I see -31.64.

Q: Is -31.64 within the range the instrument is supposed to identify the internal standard.

A: It appears to be outside.

Q: So the instrument failed to identify the internal standard.

A: That is correct.

Q: Same page looking at the F2 blank. Is that in or out of range?

A: It's outside by .02 by mil.

Q: That's for sample A, let's look at sample B. USADA 351. F3 blank, f1 fraction and f2 fraction. Let's do it one at a time. f3 blank.

A: -31.54

Q: inside or outside.

A: outside.

Q: Sample b f1.

A: -31.08

Q: inside or outside?

A: I suppose it's outside.

Q: Don't apologize, you want it done right too. When you testified earlier the QC and internal standards were impressive, did you you aware of these values?

A: Short answer, I was aware some were a bit outside, Yes.

Q: When you were talking about the cal-mix acetate, one of the impressive QC's by LNDD. Do you this at your lab?

A: No, we use something different for a similar purpose.

Q: What is it used for?

A: It's a mix of 4 steroids calibrated so there is a known isotope ratio, over the range of interest.

Q: These acetates are not in a urine matrix?

A: Correct.

Q: Is a urine matrix different?

A: Yes.

Q: The urine is more complex, right?

A: Yes.

Q: Because only what you know is in the mix?

A: Yes.

Q: Is it easier to get a clean chromatogram from mix-cal than urine?

A: Yes.

Q: Because it's designed to be pure, right?

A: Yes.

Q: Let me show you a response to a discovery request, describing the mix acetate as a positive control. What is a positive control.

A: It's a sample that is known response to test whether the instrument is giving the right response.

Q: So the purpose of the control here is to have the same isotopes as were testing for?

A: I wouldn't say so.

Q: Why not?

A: It doesn't matter what the identity is for IRMS it's not that important.

Q: If you were to have a positive control for Exo-T, shoudln't it contain the 5a and 5b?

A: My feeling is that it would not be important.

Q: It would have to elute at the same time to know right?

A: Yes.

Q: Turn to 354. These are the 4 that are added. Do these elute at the same time 5a and 5b?

A: No, they elute at the same time as metabolites of Exo-T.

Q: Precisely the same time? These 4 would elute at the same time as what we're looking for?


A: Insofar as they are the same molecules they'd elute at the same time. The 5a is not here.

Q: The 5a is not here, but it's not as the mix-cal is it?

A: No. Makes no difference.

A : The isotope ratio for any of these compounds is good enough. It's not necessary for it to be same molecule.

Q Does it need to be in a urine matrix

A: ???

Q: Because they say it's a positive control, that makes it a positive control?

A: If they call it a positive control, it is a positive control.

Q: You testified the instrument was linear.

A: Yes.

Q: This is important, correct?

A: Yes.

Q: Can you rely on it if it isn't linear?

A: Sometimes yes.

Q: How about for exo-t? If it's non-linear, would it affect the result for carbon?

A: Yes.

Q: What effect would sample concentration have on linearity?

A: None?

Q: If concentration decreases, can it affect the linearity.

A: no.

Q: You said linearity affects the accuracy of the CIR across samples.

A: yes.

Q: So linearity is critical.

A: Yes.

Q: How often do you do linearity runs at your lab.

A: One week, approximately.

Q: When measuring linearity, is it important to cover the full range of intensity expected?

A: Yes.

Q: If a machine were linear at one end of intensity, it might not be linear at another range not tested?

A: That's hypothetical. IRMS for carbon is very linear. It's a property of the material in the machine. It's not usually a problem for carbon. So what you say it hypothetically correct, it's rarely an issues.

Q: Do you test over the full range?

A: That would be reasonable.

Q: Exhibit 25, LNDD 313, 314. Do you see the smallest peak area is 26.6ns?

A: yes.

Q: Turn to LNDD 331. You see the smallest peak area is 26.9ns?

A: yes.

Q: LNDD 327. Smallest 27.98ns?

A: yes.

Q: exhibit 24, USADA 166. For endosterone, the peak area is 25.9ns

A: yes.

Q: USADA mumbme mumble

A: barely.

Q: USADA 344, 22ns?

A: Yes.

Q: USADA 355, 20.5ns.

A: Yes.

Q: Comparing the last 4, would you agree the measured peak range is not covered by the linearity test?

A: No. There's a zero. That you have measures below the lowest number they checked, but I don't believe it's relevant to the analysis.

A: Checking linearity on this test is overkill, in my opinion.

Q: You're familiar with this article?

A: Yes.

Q: This chart. Please explain to the panel.

A: Note this is a log scale chart. This the delta value determined for CO2 injected into the column. For larger amounts we get good precision, as the amount decreases, we get a "horn pattern" where the spread increases as we have smaller amounts. Test/retest precision is worse.

Q: Earlier you testified as concentration goes up and down there is a straight line.

A: As long as you have signal, you're ok.

Q: Doesn't this chart show as you reduce the amount, it matters?

A: No that's precision, not accuracy.

[Interruption for time. How much? 45minutes? Same as this morning? No, a real 45 minutes]



Craig said...

There's a comment: "[ This is part of doc pack we haven't received. ]"

Does this mean that TBV hasn't received it, or that Landis hasn't received it? Or both?

Dave and Bill: Thanks for all your excellent work, keeping those of us out in the hinterlands up to date.

Michael said...

I would think that the USADA witnesses would answer with more than, "If they call it a positive control, it is a positive control."

They aren't or don't seem to want to or really explain why Floyd tested positive. More like, the lab found him positive, so he must be positive.

I hope Floyd's experts detail WHY he didn't test positive.