It looked like 2008 would be a banner year for Sequenom, a San Diego genetic testing company.
Expectations escalated that January when the company began enrolling pregnant mothers in trials of a blood test that predicted whether their babies would have Down syndrome. A pathologist at The Chinese University of Hong Kong had developed the test and was already producing research that showed it worked.
CEO Harry Stylli boasted the test “could revolutionize Down syndrome prenatal screening,” in a press release at the time. If he was right, every pregnant woman would want to take the test, meaning huge profits for a company that had racked up a $470 million deficit.
But in the next year, those hopes became tenuous. Then they were crushed.
Sequenom’s scientists learned the test didn’t work as well as they’d hoped. They had trouble expanding the test from the research done by the pathologist, Dr. Dennis Lo, which was primarily done on Asians, to other races. When they tried to predict whether a baby would have Down syndrome, they were sometimes wrong.
But Elizabeth Dragon, the company’s vice president of research and development, didn’t admit defeat. The tests had achieved nearly 100 percent accuracy, she said, according to court documents.
“You know it’s definitely a Downs and you can read it as a Downs without any problem,” she said in a presentation in June 2008.
But that wasn’t true. The results showed success rates usually hovered around 70 to 80 percent, according to a Securities and Exchange Commission complaint.
Sequenom’s troubles quickly went from scientific to legal as a string of bad decisions pushed Dragon and her company into a downward spiral. At the end of that spiral, Sequenom found itself dealing with numerous lawsuits and investigations by the SEC, Nasdaq and the FBI for deceiving the investing public.
Quick ascents and falls are common in the biotech world, but Sequenom’s was more extreme than most. And while biotech companies are sometimes on the receiving end of lawsuits, either over patent questions, misleading stockholders, or medical liability claims, few are as publicized as Dragon’s drama.
In September 2009, Dragon, Stylli and three other employees were fired and two others resigned.
Dragon pleaded guilty to a single count of conspiracy to commit securities fraud last month, a crime that carries potential penalties of up to 25 years in prison and $250,000 in fines. She admitted to participating in “a conspiracy to defraud Sequenom’s shareholders” by presenting false information about the Down syndrome test, according to the U.S. Attorney’s Office. She acknowledged that she lied about the test’s near-perfect accuracy, and that this lie was purposefully meant to inflate Sequenom’s stock prices.
Dragon is scheduled to be sentenced in U.S. District Court on Nov. 8. The SEC also filed a separate civil complaint against Dragon that could result in a fine.
Dragon has moved to Arizona and promised the SEC to never work again as an officer or director of a public company. She has left behind the high-pressure world of for-profit science, where the complexities of the human body can make developing drugs and medical tests so difficult they conflict with commercial demands.
Now, her once-high-flying company has started from scratch, developing a new Down syndrome test based on different genetic material. And new documents released in the federal cases paint a behind-the-scenes picture of what went wrong in the weeks and months before the company’s meltdown in 2009.
♦♦♦
On a warm afternoon in September 2008, Dragon presented the most recent Down syndrome test data to analysts and investors in New York City.
She had good news to share: The test was working with 100 percent accuracy and a sample taken from a pregnant woman in her first trimester was identified correctly. That sample was especially important because the company wanted to sell a test that could be used in the first trimester, according to the SEC. Its results “are behaving exactly as you would like to see them behave,” Dragon said.
Even the fact that this kind of Down syndrome test existed was new and exciting. Sequenom’s test was based on Lo’s seminal discovery that a fetus’s genetic material could be seen in a mother’s blood.
The idea: By analyzing a blood sample from a pregnant mother, scientists could detect whether her fetus had an extra chromosome that causes Down syndrome.
But Dragon’s public display of confidence about the test hid the difficulties her team of scientists was having.
To make their predictions, the scientists looked at graphs that showed the amount of genetic material a fetus had, according to the SEC’s complaint. These graphs had two peaks representing one chromosome from the mother and one from the father.
If the fetus had Down syndrome, one peak would be twice as tall as the other, indicating it received an extra chromosome. Normal fetuses would show two peaks of the same height, meaning the fetus received one chromosome each from its mother and father.
But the results weren’t always so clear. It was sometimes hard to tell if the peaks were the same height or if one was twice as tall as the other, so the Sequenom scientists had to come up with “cutoff” points to help them decide. If the tips of both peaks landed between the upper and lower cutoff points, they called the peaks the same size, meaning the fetus was normal. Otherwise, it was labeled as having Down syndrome.
Dragon’s team ran hundreds of trials of their Down syndrome test throughout 2008 and early 2009.
When they tried to make predictions blindly based on the genetic material, they sometimes guessed wrong, according to the SEC complaint. Instead of recording these incorrect predictions, Dragon gave her scientists the actual results so they could change the cutoff points to get the right answers.
Wrong guesses and procedural corrections are not uncommon — or criminal — in science.
The problem: Instead of copping to these mixed results, Dragon lied about them, according to both the SEC and U.S. Attorney complaints. In addition to her presentation on that warm September afternoon, she gave two other talks to analysts and investors, who would be deciding whether to buy Sequenom’s stock, in which she said the tests were run blindly and achieved almost 100 percent accuracy.
“The Sequenom R&D scientists did not know the results before they were [tested], so it was blinded,” she said in January. “That was very important for us.”
As Dragon shared these results, interest in her company climbed higher. Between her first presentation on June 3, 2008 and her last at the end of January 2009, Sequenom’s stock price rose from $7.66 to $22.60.
The highpoint: $27.76. It came the day after her presentation to analysts and investors in New York City.
That year Sequenom’s board of directors decided that based on the company’s achievements, Dragon deserved a $58,618 bonus of cash and stock on top of her $279,133 salary.
As the head of Sequenom, Stylli also had a lot to gain. He earned an $180,810 bonus and the esteem that comes with turning around a faltering company.
“Life is good at Sequenom these days,” Luke Timmerman wrote in Xconomy.com in October 2008. “CEO Harry Stylli was so relaxed when I stopped by for an interview last week, he was practically lying down in one of the guest chairs in his office at one point while telling me the company’s story.”
On April 9, 2009, Stylli trumpeted the test’s results in the annual report to shareholders, saying it “produced remarkable results.”
Twenty days later, the company spoke with a very different tone. On April 29, Sequenom released a statement announcing that the launch of their Down syndrome test would be delayed “due to the discovery of employee mishandling of research and development test data and results.”
It’s not known how the discovery came to light. Spokesman Ian Clements said Sequenom would only reveal that it happened internally.
The test had been expected by that June, but instead the statement promised it would be out by the end of the year.
♦♦♦
The company called the revelation “a temporary setback” and said that its Down syndrome test was still scientifically sound.
But a report to the company’s shareholders sounded slightly less optimistic, saying the announcement severely hurt its reputation, “which is likely to take a significant amount of time and effort to repair.”
Overnight, the price of Sequenom’s stock fell from $14.91 to $4.69, and in the months after the announcement, the company found itself with several messes to clean. Stockholders filed lawsuits, the San Diego-based TrovaGene sued because it had licensed technology to Sequenom and claimed it suffered damages from the Down syndrome test fallout, not to mention the investigations.
In an attempt to clean up, three members of Sequenom’s board formed an independent committee to look into what happened and hired a law firm to run the investigation.
The committee came back in September 2009 with a list of recommendations that included firing Dragon, Stylli and three unnamed scientists from Dragon’s team. The chief financial officer and the vice president for commercial development of prenatal diagnostic tests also resigned.
Sequenom’s board also promised to better train employees in ethics and scientific processes, increase use of independent third parties in research and use new procedures for the storage and management of samples for testing.
For example, when the company begins testing its new Down syndrome test this fall, it will partner with academic institutions who will independently run trials and publish those results in peer-reviewed scientific journals, Clements said.
In May, Sequenom agreed to pay stockholders $14 million, their attorneys $2.5 million and give stockholders 6.8 million shares of common stock. TrovaGene’s lawsuit, which sought $300 million in damages, was dismissed because it was filed in New York, not San Diego, but it may rise up again.
Clements said that he won’t comment on any of the investigations, but that Sequenom is fully cooperating with the FBI, Nasdaq and the SEC.
After everything that has unfolded, a simple statement to Sequenom’s stockholders on April 28 might summarize the Down syndrome test events best:
“2009 was a challenging year.”
Contact Claire Trageser directly at claire.trageser@gmail.com. And follow her on Twitter: @clairetrageser.