February 27, 2012
Why Victor Conte’s Idea Doesn’t Work

Victor Conte is an interesting guy. I find him intriguing and engaging … and knowledgeable. I mean, the guy was one of the brains behind BALCO. 

But when it comes to one aspect of drug testing — one that is coming up again with the Ryan Braun decision — he’s dead wrong. Well, not so much wrong as unrealistic. Victor has called for the use of carbon isotope ratio testing on all drug samples. (Let’s ignore that isotope ratio mass spectrometry is considered the state of the art in this discussion.) 

Currently, testers in MLB do a quick test - a ratio of testosterone to epitestosterone - to see whether or not a test is likely to contain added testosterone. Until the mid-2000’s this was the ONLY test used. It is cheap, relatively effective, but is also very easy to beat* and has a false positive rate that’s too high compared to modern testing sensitivities. 

(* Conte and BALCO did exactly this. “The Cream” was merely a compounded version of prescription Androgel mixed to keep the T/E ratio within the standards of the time. Androgel and other synthetic testosterone prescriptions are still available. Axiron, an underarm application of this gel, is currently in heavy rotation.) 

So why is the ratio test done? Its very simple. Cost and time. The ratio test takes a matter of moments and can be largely automated. The IRMS (or CIR, if you insist, Victor) is both timely and expensive. MLB does well over 3000 samples in a year, but lets use that for a nice round number and easy math. According to a local lab that performs these types of tests for corporate and government clients, this type of test would cost a couple hundred dollars (depending on volume discounts) for each test. The bigger problem would be that the test takes about an hour each to be prepared, sampled, analyzed, and then to clean the machine, which is more key than most realize. 

At 3000 samples, you’re talking about 3000 hours or about 125 man/days. That’s over a full year of eight hour days doing nothing but MLB samples (and a low end estimate on numbers as well.) That means that a lab could do nothing but that using the IRMS, necessitating the purchase of a new piece of equipment for a single client. That’s going to be very costly in every meaning of the word. 

You might say “The Olympics do it.” It’s true that the Olympics do have a large number of participants (over 11,000 at the Beijing games) but they tend not to test all their participants, focusing instead on medal winners and random testing. Even then, the Olympics use multiple labs and take months to fully test all samples. 

The other issue is that IRMS is a destructive test. The urine is taken out of the collection device, placed in a testing vessel, and burned as part of the test. You don’t get a second chance at it. If ONLY the IRMS were done, the current collection amounts would work, but there would be no way to re-test unless a “B” sample were held and stored, which we already know is a problem. Doing two IRMS tests doesn’t prove much and only adds to the issue of workload.

Conte is not wrong. The IRMS (or CIR) tests are more sensitive and “better.” They’re also far more expensive and create a workload that is untenable, even for the world-class lab baseball currently uses. I’m sure Thermo Electron, the company that has a near-monopoly on this type of testing equipment, would love to sell a few more machines though you can imagine both the cost and the backlog. 

For Conte’s idea to work, someone is going to have to invent a newer, faster, cheaper method of testing … or they’re going to need a giant pile of cash.

UPDATE: After posting this, I had a Twitter conversation with Victor Conte. Victor asserted that it was possible to do this and did some checking. Victor says that an IRMS machine can handle 400 tests per month. That’s only slightly higher than the numbers I was given. Again, it’s essentially taking one machine and making it purely MLB’s testing machine, which adds costs. The Montreal lab has more than one client that uses an IRMS machine and while Conte’s point that there’s more than one lab, in MLB’s case, there’s only one they use.

So, I will agree — and never asserted otherwise — that Conte’s numbers show this type of testing to be possible, I still disagree that it is plausible. The cost in machinery and man-hours remains excessive for the increase in accuracy. How much cost? An IRMS machine costs well over a million dollars and is really only available from one company (see above.) In addition to the technicians — you’d need several — there is maintenance, additional storage, and floor space for these large, technical, and fragile machines and their supplies. On a per-year basis, the testing budget to allow this kind of procedure would triple what MLB spends. 

UPDATE TWO: After publishing the update, Conte and I continued discussing this. In the course of that, I suggested a survey approach. That is, MLB would take a “significant sample” - let’s say 500 samples - and run them through IRMS without the trigger. Conte believes we would find evidence of micro dosing and/or people beating the ratio. I’d definitely be curious. This approach was used in 2003 and if it found a significant problem, I would agree that the costs should be bourne. I hope the NFL is listening as well.

  1. willcarroll posted this