Asked if he believed Armstrong "was using EPO based on his hematocrit levels," Swart testified in his deposition: "That's the only way you could come to that conclusion. There's no other way your hematocrit would be that high."
Armstrong denied he was using EPO. Under questioning at the arbitration hearing, he was asked:
"Do you recall if any riders for the Motorola team tested with hematocrit levels above 50?"
"No ... Certainly not myself," Armstrong responded.
There was no test for EPO at that time, nor in 1999 when Armstrong broke through to win his first Tour de France. Scientists developed testing methods that were first implemented in professional cycling in 2001. The tests have always required scientific interpretations that continue to make them controversial.
As part of a project in 2004 to refine lab procedures, researchers used archived urine samples from earlier years. That's how a collection of Armstrong's 1999 race samples happened to be tested years later.
The results, Australian researcher Michael Ashenden testified in Dallas, show Armstrong's levels rising and falling, consistent with a series of injections during the Tour. Ashenden, a paid expert retained by SCA Promotions, told arbitrators the results painted a "compelling picture" that the world's most famous cyclist "used EPO in the '99 Tour."
The tests were conducted on what doping authorities call the "B sample," the second half of urine samples traditionally used in doping tests. The "A samples" were used and discarded in 1999.
Armstrong's B samples -- actually a collection of several samples provided at various points throughout the 1999 race -- were among samples of several supposedly anonymous racers taken out of deepfreeze in 2004.
The process is supposed to involve lab control numbers only, not an athlete's name. However, a French newspaper matched up Armstrong's name to at least six positive samples using race documents obtained from the International Cycling Union.
To make a typical doping case, then and now, authorities must prove a positive A and B sample. Such conclusive testing is not possible in the absence of an A sample.
Nonetheless, Ashenden testified that results he analyzed showed a "patent abuse which very, very closely resembles what I would suspect to see in an athlete actually using EPO."
Ashenden is a noted expert on blood doping. He coordinated research resulting in a two-year suspension of another former Armstrong teammate, Tyler Hamilton, found to be transfusing with someone else's blood in 2004.
The EPO test separates proteins by their electrical charge -- that is, when a sample is run through an electrically charged gel, it separates those produced naturally in the body from those of injected EPO. Scientists then read the intensities and positioning of natural and injected EPO isoforms, or molecules, on the gel.
Any reading over 80% traditionally was viewed as positive for the presence of injectable EPO.
Ashenden provided arbitrators with a day-by-day breakdown of Armstrong's test results from the 1999 samples. For example:
On July 3, 1999, Armstrong won the first race of the Tour -- the 4.2-mile prologue. His doping control form shows he was tested at 9:45 a.m. Ashenden said Armstrong's reading was 100%.
Such a high level, Ashenden testified, "is consistent with an injection that was received within just a few hours."
July 13, the first day in the Alps, ended in Sestriere, Italy. Armstrong took a six-minute lead. He was tested at 5:15 p.m. His test reading: 96.6%, consistent with an injection he "would have received -- could have received earlier in ... the day," Ashenden testified.
Tests for the final six racing stages showed "there was never enough EPO," natural or otherwise, "in any of Armstrong's urine samples to report a result," Ashenden said.
His explanation: When an athlete takes injectable EPO, the levels of that injectable EPO fall off day by day. At the same time, the kidneys have stopped producing natural EPO because the body recognizes "there's too much blood in his circulation."
The result, he testified, is that there isn't enough EPO of any sort to measure as the body gets a "chance to come to its ... natural level." The Australian researcher concluded that "not finding enough EPO in the sample to analyze ... you see that when an athlete stops taking EPO injections."
Ashenden defended the testing process, saying researchers had no way of knowing whose samples they were testing at the time.
A French newspaper reported last fall that three other racers tested positive for EPO during the same trials -- cyclists from Spain, Denmark and Colombia. All three cyclists later denied using EPO.
Armstrong's lawyers tried to block Ashenden's testimony, arguing it was unfair and "unduly prejudicial." The arbitration panel denied that bid.
Armstrong flatly rejected Ashenden's analysis. He called the claims a "pure witch hunt."
"When I gave the sample, there was no EPO in the urine," he testified at the hearings.