By ALEX BERENSON
Published: December 17, 2006
The drug maker Eli Lilly has engaged in a decade-long effort to play down the health risks of Zyprexa, its best-selling medication for schizophrenia, according to hundreds of internal Lilly documents and e-mail messages among top company managers.
The documents, given to The Times by a lawyer representing mentally ill patients, show that Lilly executives kept important information from doctors about Zyprexa’s links to obesity and its tendency to raise blood sugar — both known risk factors for diabetes.
Lilly’s own published data, which it told its sales representatives to play down in conversations with doctors, has shown that 30 percent of patients taking Zyprexa gain 22 pounds or more after a year on the drug, and some patients have reported gaining 100 pounds or more. But Lilly was concerned that Zyprexa’s sales would be hurt if the company was more forthright about the fact that the drug might cause unmanageable weight gain or diabetes, according to the documents, which cover the period 1995 to 2004.
Zyprexa has become by far Lilly’s best-selling product, with sales of $4.2 billion last year, when about two million people worldwide took the drug.
Critics, including the American Diabetes Association, have argued that Zyprexa, introduced in 1996, is more likely to cause diabetes than other widely used schizophrenia drugs. Lilly has consistently denied such a link, and did so again on Friday in a written response to questions about the documents. The company defended Zyprexa’s safety, and said the documents had been taken out of context.
But as early as 1999, the documents show that Lilly worried that side effects from Zyprexa, whose chemical name is olanzapine, would hurt sales.
“Olanzapine-associated weight gain and possible hyperglycemia is a major threat to the long-term success of this critically important molecule,” Dr. Alan Breier wrote in a November 1999 e-mail message to two-dozen Lilly employees that announced the formation of an “executive steering committee for olanzapine-associated weight changes and hyperglycemia.” Hyperglycemia is high blood sugar.
At the time Dr. Breier, who is now Lilly’s chief medical officer, was the chief scientist on the Zyprexa program.
In 2000, a group of diabetes doctors that Lilly had retained to consider potential links between Zyprexa and diabetes warned the company that “unless we come clean on this, it could get much more serious than we might anticipate,” according to an e-mail message from one Lilly manager to another.
And in that year and 2001, the documents show, Lilly’s own marketing research found that psychiatrists were consistently saying that many more of their patients developed high blood sugar or diabetes while taking Zyprexa than other antipsychotic drugs.
The documents were collected as part of lawsuits on behalf of mentally ill patients against the company. Last year, Lilly agreed to pay $750 million to settle suits by 8,000 people who claimed they developed diabetes or other medical problems after taking Zyprexa. Thousands more suits against the company are pending.
On Friday, in its written response, Lilly said that it believed that Zyprexa remained an important treatment for patients with schizophrenia and bipolar disorder. The company said it had given the Food and Drug Administration all its data from clinical trials and reports of adverse events, as it is legally required to do. Lilly also said it shared data from literature reviews and large studies of Zyprexa’s real-world use.
“In summary, there is no scientific evidence establishing that Zyprexa causes diabetes,” the company said.
Lilly also said the documents should not have been made public because they might “cause unwarranted fear among patients that will cause them to stop taking their medication.”
As did similar documents disclosed by the drug maker Merck last year in response to lawsuits over its painkiller Vioxx, the Lilly documents offer an inside look at how a company marketed a drug while seeking to play down its side effects. Lilly, based in Indianapolis, is the sixth-largest American drug maker, with $14 billion in revenue last year.
The documents — which include e-mail, marketing material, sales projections and scientific reports — are replete with references to Zyprexa’s importance to Lilly’s future and the need to keep concerns about diabetes and obesity from hurting sales. But that effort became increasingly difficult as doctors saw Zyprexa’s side effects, the documents show.
In 2002, for example, Lilly rejected plans to give psychiatrists guidance about how to treat diabetes, worrying that doing so would tarnish Zyprexa’s reputation. “Although M.D.’s like objective, educational materials, having our reps provide some with diabetes would further build its association to Zyprexa,” a Lilly manager wrote in a March 2002 e-mail message.
But Lilly did expand its marketing to primary care physicians, who its internal studies showed were less aware of Zyprexa’s side effects. Lilly sales material encouraged representatives to promote Zyprexa as a “safe, gentle psychotropic” suitable for people with mild mental illness.
Some top psychiatrists say that Zyprexa will continue to be widely used despite its side effects, because it works better than most other antipsychotic medicines in severely ill patients. But others say that Zyprexa appears no more effective overall than other medicines.
And some doctors who specialize in diabetes care dispute Lilly’s assertion that Zyprexa does not cause more cases of diabetes than other psychiatric drugs. “When somebody gains weight, they need more insulin, they become more insulin resistant,” Dr. Joel Zonszein, the director of the clinical diabetes center at Montefiore Medical Center in the Bronx, said when asked about the drug.
In 2003, after reviewing data provided by Lilly and other drug makers, the F.D.A. said that the current class of antipsychotic drugs may cause high blood sugar. It did not specifically single out Zyprexa, nor did it say that the drugs had been proven to cause diabetes.
The drugs are known as atypical antipsychotics and include Johnson & Johnson’s Risperdal and AstraZeneca’s Seroquel. When they were introduced in the mid-1990s, psychiatrists hoped they would relieve mental illness without the tremors and facial twitches associated with older drugs. But the new drugs have not proven significantly better and have their own side effects, said Dr. Jeffrey Lieberman, the lead investigator on a federally sponsored clinical trial that compared Zyprexa and other new drugs with one older one.
The Zyprexa documents were provided to the Times by James B. Gottstein, a lawyer who represents mentally ill patients and has sued the state of Alaska over its efforts to force patients to take psychiatric medicines against their will. Mr. Gottstein said the information in the documents raised public health issues.
“Patients should be told the truth about drugs like Zyprexa,” Mr. Gottstein said.
Lilly originally provided the documents, under seal, to plaintiffs lawyers who sued the company claiming their clients developed diabetes from taking Zyprexa. Mr. Gottstein, who is not subject to the confidentiality agreement that covers the product liability suits, subpoenaed the documents in early December from a person involved in the suits.
In its statement, Lilly called the release of the documents “illegal.” The company said it could not comment on specific documents because of the continuing product liability suits.
In some ways, the Zyprexa documents are reminiscent of those produced in litigation over Vioxx, which Merck stopped selling in 2004 after a clinical trial proved it caused heart problems. They treat very different conditions, but Zyprexa and Vioxx are not entirely dissimilar. Both were thought to be safer than older and cheaper drugs, becoming bestsellers as a result, but turned out to have serious side effects.
After being pressed by doctors and regulators, Merck eventually did test Vioxx’s cardiovascular risks and withdrew the drug after finding that Vioxx increased heart attacks and strokes.
Lilly has never conducted a clinical trial to determine exactly how much Zyprexa raises patients’ diabetes risks. But scientists say conducting such a study would be exceedingly difficult, because diabetes takes years to develop, and it can be hard to keep mentally ill patients enrolled in a clinical trial.
When it was introduced, Zyprexa was the third and most heralded of the atypical antipsychotics. With psychiatrists eager for new treatments for schizophrenia, bipolar disorder, and dementia, Zyprexa’s sales soared.
But as sales grew, reports rolled in to Lilly and drug regulators that the medicine caused massive weight gain in many patients and was associated with diabetes. For example, a California doctor reported that 8 of his 35 patients on Zyprexa had developed high blood sugar, including two who required hospitalization.
The documents show that Lilly encouraged its sales representatives to play down those effects when talking to doctors. In one 1998 presentation, for example, Lilly said its salespeople should be told, “Don’t introduce the issue!!!” Meanwhile, the company researched combinations of Zyprexa with several other drugs, hoping to alleviate the weight gain. But the combinations failed.
To reassure doctors, Lilly also publicly said that when it followed up with patients who had taken Zyprexa in a clinical trial for three years, it found that weight gain appeared to plateau after about nine months. But the company did not discuss a far less reassuring finding in early 1999, disclosed in the documents, that blood sugar levels in the patients increased steadily for three years.
In 2000 and 2001, more warning signs emerged, the documents show. In four surveys conducted by Lilly’s marketing department, the company found that 70 percent of psychiatrists polled had seen at least one of their patients develop high blood sugar or diabetes while taking Zyprexa, compared with about 20 percent for Risperdal or Seroquel. Lilly never disclosed those findings.
By mid-2003, Lilly began to change its stance somewhat, publicly acknowledging that Zyprexa can cause severe obesity. Marketing documents make clear that by then Lilly believed it had no choice. On June 23, 2003, an internal committee reported that Zyprexa sales were “below plan” and that doctors were “switching/avoiding Zyprexa.”
Since then, Lilly has acknowledged Zyprexa’s effect on weight but has argued that it does not necessarily correlate to diabetes. But Zyprexa’s share of antipsychotic drug prescriptions is falling, and some psychiatrists say they no longer believe the information Lilly offers.
“From my personal experience, at first my concerns about weight gain with this drug were very significantly downplayed by their field representatives,” said Dr. James Phelps, a psychiatrist in Corvallis, Or. ‘Their continued efforts to downplay that, I think in retrospect, was an embarrassment to the company.”
Dr. Phelps says that he tries to avoid Zyprexa because of its side effects but sometimes still prescribes it, especially when patients are acutely psychotic and considering suicide, because it works faster than other medicines.
“I wind up using it as an emergency medicine, where it’s superb,” he said. “But I’m trying to get my patients off of Zyprexa, not put them on.”