Under Pressure

There are a few agencies under pressure these days, like USADA who are under pressure to get the Landis case heard and settled. Very likely that’s why USADA is beginning to be more cooperative in the last couple of weeks. WADA wants the case settled, the UCI wants the case settled, and the Tour organizers want the case settled, among others. And the AFLD, France’s anti-doping authority, is putting the pressure on USADA, too.

But this post is about pressure of a different kind. The kind used (or more accurately measured) in mass spectrometry. As TBV reported on Sunday, Arnie Baker has made comments that some (or perhaps all) of the mass spectrometry tests performed on Landis’ samples were done at a pressure higher than the manufacturer of the MS instrument recommends.

At yesterday’s “Tour of Innocence” event (as the newspapers have taken to calling these meetings) in San Jose, more details emerged. TBV went to this meeting too and took a camera to photograph some of the new slides. One of them (note: you may need to scroll around the photo a bit to see the meaty parts of the information) contains the information on Baker’s comments about pressure readings outside the manufacturer’s specification.

Apparently, the pressure data came from a lab documentation pack provided to the Landis defense team from AFLD. For whatever reason, USADA’s documentation doesn’t include the data.

This information sheds a whole lot of light on what problem Baker refers to in his latest slideshow. Hopefully, this explanation won’t get too complicated. I’d hate to contradict Peloton Jim’s comment that this is “physics for poets” and all. 😉

In my previous posts, I was operating under the belief that Baker was commenting on the internal pressure of the system being too high. Well, he is, sort of. But not in the way that I expected. And the implications are different than what I expected. But they’re as or more important that what I’ve believed up to now.

What’s going on with the pressure gauge that Baker’s documentation refers to is measurement of a vacuum. The pressures being measured are very small. Less than one-thousandth of normal air pressure at sea level. These pressures need to be low to ensure that no contamination occurs. From the slide TBV photographed, we can see the following:

3. Wait until the pressure shown on the Penning gauge falls below 5E-6 mbar. If there are no major leaks along the inlet capillaries the pressure will fall quickly and settle to the operating pressure between 2 and 4E-6 mbar. Failure to reach the operating pressure indicates major leaks. These must be cured before proceeding any further.

Alongside this information is an inset of some lab documetation showing the operating pressure at the Penning gauge to be 5.2E-6 mbar, which exceeds the manufacturer’s recommended operating range. It’s not clear which test this information comes from, and whether this is consistent throughout all of the tests conducted (A sample, B sample, etc.).

The manufacturer’s documentation says that if the pressure exceeds a certain level (5/1000000 millibars) then the system has leaks. Now leaks can work two ways, something entering the system or something exiting the system. In either case, it’s bad. But for a mass spectrometer, a leak near an inlet capillary indicates air or gas entering the system from somewhere that it’s not supposed to.

And that presents a huge problem, as we’ll see in a moment. But first, let’s revisit what IRMS is and how it’s used in anti-doping cases. The idea behind IRMS is that it’s measuring the ratio of one isotope of an element to another isotope. In the case of anti-doping tests, they’re interested in the ratio of Carbon¹³ to Carbon¹², which they convert into a percentage.

The conjecture is that any ratio that exceeds a certain percentage is proof positive of doping. I say conjecture purposely. I am not convinced that this is a demonstrated “scientific fact,” as WADA has their own research that shows it’s possible through diet to exceed their own threshold. Interestingly, ScienceDaily published an article recently about WADA’s approach to “science.”

For the Landis case, or any anti-doping case, the science is assumed (by WADA and the anti-doping process) to be infallible. So Landis’ defense team must show, without challenging the basic science, why the results are wrong. And here is where our leaky friend, the mass spectrometer, comes in to play.

How IRMS works, in a nutshell, is this. The instrument separates the compound being studied into electrically charged components called “ions” and then separates the ions based on their mass to charge ratio. To get an accurate reading of what you’re studying, you need to ensure that your sample is pure. If it isn’t, you’ll be measuring ions from contaminants, which may influence your results.

So what would happen if you were measuring contaminants mixed in with the sample while running an IRMS study like the ones performed in the Landis case? Well, that depends on what contaminants enter the system. If, for example, they were different isotopes of Carbon, you could get a different result for the ratio of Carbon¹³ to Carbon¹².

Whether or not Landis’ sample became contaminated during testing (and during which testing) is hard to determine from the information available. However, the machine was being operated outside the manufacturer’s recommendations, and the manufacturer’s documentation says specifically that the machine must reach a certain operating pressure in order to proceed. Implicit in their documentation is that to proceed when the pressure was above that range would yield invalid results. LNDD’s operation of the mass spectrometer makes their data problematic as, according to the manufacturer’s documentation, it had “major leaks” around the inlet capillaries which should have been corrected.

LNDD’s failure to have the proper documentation and failure to operate their instruments properly is completely inexcusable and almost certainly a violation of any certification of accreditation the lab has obtained. It also means that the results that LNDD found can’t be considered reliable, as there is a very real possibility that contamination may have occurred during the testing process. The manufacturer clearly states that leaks must be “cured” given the pressure readings for the LNDD instrument.

Let’s put it another way: Because of this development, all of the data LNDD gathered using their mass spectrometer at the pressure shown in Baker’s slideshow has to be considered invalid. If this were true for all of the testing done — the A sample tests, as well as the B sample tests — then every bit of data in the Landis case is invalid. And if that’s what happened, then this case should be dropped. Now.