MEMPHIS, Tenn. — Dr. Gary Murray, a cardiologist, was faced with a common medical dilemma: He had no idea what was wrong with his patient.
Milton Cole, a 71-year-old man in generally good health, was complaining of chest pains. A battery of cardiac tests couldn't pinpoint the problem. To blunt the pain, Murray gave his patient a prescription and some free samples of the drug.
The drug Murray prescribed was Prozac, a popular antidepressant that isn't approved by the Food and Drug Administration for treating chest pain. Murray later said he had no idea that experts had debated for years whether Prozac caused suicide.
Thirteen days after that visit to the doctor, on June 28, 2001, Cole's wife, Amby, found him hanging from a beam in a back room of their shop.
"This was a patient of mine and I was trying to help him," Murray recalled. "I'm completely upset. I'll be that way forever."
That a heart specialist even had free supplies of a drug that's usually the province of psychiatrists says a lot about how drug marketing today encourages physicians to prescribe medications for unapproved, or off-label, uses.
By offering specialty drugs to non-specialists, sending salesmen to doctors' offices and medical conventions, and touting their drugs' benefits on the slimmest of evidence, pharmaceutical companies have sent off-label retail sales soaring.
Off-label sales of the top-selling drugs Knight Ridder studied hit $12.9 billion in the last year, producing nearly a quarter of those drugs' retail sales.
With an aging population, a shift to drug-based health care and the prospect of a massive government prescription-insurance plan, prescriptions for unapproved uses are only likely to accelerate.
Promoting this growth is a symbiotic relationship between physician and drug makers in which sales representatives routinely target doctors untrained in the basics of drug therapy and with little time, inclination or independent information to assess a medication's usefulness or its risks.
Consider this courtroom conversation involving a lawyer representing Schering-Plough Corp., the mammoth U.S. drug maker.
A quizzical U.S. District Judge Lawrence McKenna of New York asked whether doctors "in any significant number" really based their opinions about drugs on pitches from sales reps.
"Yes, your honor," said Gregory Diskant, Schering-Plough's outside attorney.
"That's scary, isn't it?" the judge asked.
"It is scary. It is scary, but it is true," Diskant said. "You know what the truth is? (Sales calls) are the quickest, easiest way for the doctor to learn about the drug. It is a dangerous, largely unregulated phenomena in the wrong hands."
The day Cole came in for his checkup, Murray had a drug closet stocked with dozens of medicines that cardiologists commonly prescribe, as well as some that general practitioners and other specialists use.
With little to go on beyond Cole's complaints of chest pain, Murray said, he had a choice between doing nothing and attempting to ease his patient's pain.
"I chose Prozac probably because I had samples of it," he said in an interview. "I thought it was a pretty harmless thing to do."
The FDA has approved Prozac to treat depression, panic, obsessive-compulsive disorder and an eating disorder, but not pain. Murray didn't think Cole was depressed. The Brighton, Tenn., resident was upbeat, busy with friends and church, preparing for a new grandchild and a new house.
Doctors have prescribed antidepressants off-label for years to manage chronic pain, and as far as Murray knew, Prozac didn't have any serious side effects. He said he had never heard it might cause suicide. Articles debating the issue were published primarily in psychiatric journals throughout the 1990s.
A Knight Ridder analysis of government data found that over the last decade, 40 percent of Prozac prescriptions were written by nonpsychiatrists. Over the past year, 500,000 Prozac prescriptions were for off-label uses.
Prozac didn't help Cole's chest pain.
Soon, he complained of feeling jittery. His fingers tingled; he became easily aggravated. Days later, he hanged himself.
Amby Cole recounted the day she found him, describing how she rubbed her dead husband's neck where the rope had hurt him. "It was just horrible," she said, her hands trembling.
She blames drug maker Eli Lilly for not warning that Prozac can cause suicide, and her lawyer, Andy Vickery of Houston, has accused Lilly of overpromoting Prozac to nonpsychiatrists. Lilly settled the case earlier this year. The amount is confidential.
Vickery said it was his third settlement with Lilly over Prozac-suicide cases and off-label uses. Patients in the other two cases were given Prozac to treat migraines and for Tourette's syndrome, a neurological condition that produces uncontrollable tics.
Lilly officials said they had settled some lawsuits for economic reasons, but wouldn't comment on specifics. A spokeswoman, Tarra Ryker, said Lilly "does not condone or encourage off-label use of any of our medications, including Prozac."
The company long has contended that depression causes suicide, not Prozac. Even so, U.S. and British regulators warned last summer that Paxil, a drug in the same class as Prozac, may increase the suicide risk for children and adolescents, thus reviving the debate that has simmered for a decade. Ryker said this had no bearing on Prozac.
Murray still prescribes Prozac for pain, but now he warns patients of possible psychiatric side effects.
Federal law prohibits drug makers from advertising or promoting off-label drug uses, and since 1998 the FDA has cited companies nearly 70 times for improperly promoting their drugs that way, a review of its records shows.
Yet off-label promotions are commonplace, as was alleged in a recent whistleblower lawsuit against a company that's now part of Pfizer Inc. A former employee of the drug company said it employed a range of tactics to boost off-label sales of the epilepsy drug Neurontin. The widely reported case is pending.
Getting around FDA marketing rules isn't difficult. The agency acknowledges that it's impossible to police the millions of conversations between drug companies and doctors.
At the annual conference of the American Academy of Pain Management this summer at the Adam's Mark hotel in downtown Denver, 75 makers of drugs, medical devices and dietary supplements staffed sales booths, beckoning doctors with trinkets, slick promotional brochures, medical journal articles, even stopwatches.
At the Merck & Co. booth, Connie Mack said Vioxx, the blockbuster arthritis drug, was approved for certain types of pain. Then without prompting, Mack added: "They are using it pre-emptively, too," such as before surgery, she said. No, it's not approved that way, she said, but Merck could send some additional information. "They actually have a whole database now on pre-emptive use," she said.
At the Allergan Inc. booth, a representative described an approved use for Botox. She then volunteered, "There are physicians who have been using it off-label for lower back pain and migraine pain. That's all off-label." Asked about research to support such uses, she said, "there are no major gold-standard studies, but the physician community has been using it. They're telling us it is effective." She offered to send more information.
One of the biggest booths was for Cephalon Inc., based outside Philadelphia. Two big signs pitched Gabitril, a drug approved only to treat seizures. Four sales representatives in blue blazers answered questions.
Why, one was asked, was an epilepsy drug being hawked at a conference for pain doctors? "Most of the anti-epileptics have multiple other uses," replied Cephalon's Janeen Morgan.
Cephalon relies more on off-label sales than other companies. A Wall Street analyst touted "significant opportunities for off-label sales" as a reason to be bullish on the company, and Cephalon at times has aggressively marketed its drugs off-label; a 2002 letter from the FDA cited the company for making "misleading claims" about a "variety of unapproved uses" for another of its drugs, Provigil.
The company's messages apparently got through to doctors.
In Maine, for example, state Medicaid officials in 2002 noticed a growing amount of Provigil use.
Although approved at the time only for patients with narcolepsy, a disorder associated with feelings of pronounced sleepiness, Medicaid officials were receiving claims for its use to treat multiple sclerosis fatigue, attention deficit disorder, depression and "miscellaneous fatigue."
The Medicaid claims topped $1 million, with one doctor responsible for $370,000 of them, state records show.
From 2000 to 2003, 60 percent of Cephalon's sales were of Gabitril or Provigil, two drugs for which the majority of the written prescriptions are off-label. According to the prescription data that Knight Ridder analyzed, 88 percent of Gabitril's retail sales were off-label over the last year.
Cephalon has told the Securities and Exchange Commission that the "market for the approved indications of two of our three largest products is relatively small."
Even as the company begins to study whether Gabitril works for pain, tens of thousands of prescriptions for that use were filled in the last year, Knight Ridder's analysis showed.
Michael Fielder of Kansas City, Mo., said he'd taken Gabitril for the last 18 months to curb the pain of sickle cell anemia, a debilitating blood disorder. He got relief only after his doctor added morphine to the mix.
When asked for evidence that their drug may be useful for pain, Cephalon officials cited five studies — four of them with 10 or fewer patients.
Fielder is still taking Gabitril, because, he said, his doctor believes it helps. Over the years, doctors have given him Neurontin, another epilepsy drug, and Paxil, the antidepressant, for pain. None of them worked well, he said.
"When they try out something new, sometimes I don't think the doctors know if what they are using works," Fielder, 27, said recently.
Court records contain dozens of comments by some of the government's top health experts about doctors and prescription drugs. Many of them aren't flattering. The assessments about off-label prescribing are worse.
Robert J. Temple, a top official in the FDA's drug division, said in a court deposition in 1996: "I certainly believe in their good faith. I think they are trying. I don't believe that they necessarily have a capacity to get it right."
The FDA's lawyers said in a 1998 brief: "While physicians may believe that they are in a better position than FDA to evaluate off-label claims, both the evidence and the law say otherwise. Physicians tend to have confidence in their own ability to critically assess off-label information. The studies demonstrate, however, that such confidence is often unwarranted and incorrect."
Michael Wilkes, the vice dean of the medical school at the University of California, Davis, doesn't think that's surprising.
"I think it is embarrassing that so much of our practice is prescribing drugs and it's a joke how little our students and residents know about pharmacology," he said. "And once you graduate, how does a doctor learn about new medicines? It's from the pharmaceutical companies."
Wilkes, who studied the issue of off-label promotion for the FDA in the late 1990s, pointed to two rigorous studies that concluded that doctors — despite their protestations to the contrary — are swayed by pharmaceutical promotions.
"What gets marketed hard is what gets prescribed," said Jay Cohen, an adjunct associate professor of family and preventative medicine at the University of California-San Diego.
The marketing is massive.
Over the last decade, the number of drug company sales reps has more than doubled to 94,000, one for every seven doctors in the nation. In 2002, the value of the free drug samples passed out to physicians reached $11.9 billion, up more than 140 percent since 1996.
At the same time, pharmaceutical companies' spending for advertising in the traditional medical journals fell, according to IMS Health, a medical-data company.
In addition, there's evidence that when the FDA tries to get doctors' attention, they pay little heed.
Since 1990, 16 drugs have been pulled from the market for safety reasons. A review of FDA and other records shows that 11 of those were yanked, in part, because physicians didn't follow label instructions or disregarded the FDA's warnings.
Nonetheless, many physicians continue to think their practices are the best places to assess the worth and risks of prescription drugs.
Dr. Wasim Niazi, a neurologist from Rockledge, Fla., is a fan of Topamax, an epilepsy drug that's widely prescribed for conditions the FDA didn't approve it to treat. Niazi said Topamax had become the most-prescribed drug in his practice. He said he had hundreds of patients on it and wrote 10 new prescriptions a day.
While he also prescribes it for epilepsy, Niazi said, he mostly uses it to treat pain and migraines. He's also given it for tremors and to assist with weight loss. He considers it effective for all these uses, if his patient can tolerate the side effects.
Niazi said the basis for his prescribing of Topamax was his own clinical experience and what he'd learned from other doctors. What studies say about its usefulness means little to him. "Most of the literature is garbage driven by the economics," he said. And he's dismissive of clinical trials that are contrary to his own observations. "The real world is different than trials."
At first glance, the results of a clinical study of Evista, an Eli Lilly osteoporosis drug, looked promising. Among other things, researchers noticed slightly fewer breast cancers among women who took Evista versus those who were given sugar pills.
As baby boomers age, that could mean the makings of a blockbuster drug.
Unfortunately for Lilly, cancer experts didn't think the numbers meant much.
On May 18, 1998, Eli Lilly convened a focus group of doctors who were attending the American Society of Clinical Oncology annual meeting. Shown the Evista breast cancer data, they responded that the study was too small, the time frame too short and the risk of patients having strokes too great.
Promoting the drug to prevent breast cancer would be "an egregious stretch," one doctor said. It also was called an attempt to "cash in on a byproduct of the study." A proper study to measure breast cancer prevention "should last at least 10 years, preferably 20," doctors said, according to the company's meeting notes.
Even so, the cancer physicians feared that primary-care and women's-health doctors would prescribe Evista for breast cancer prevention anyway.
A month later, Lilly conducted a different survey, this one of 341 primary-care physicians and women's-health doctors. The company tested various advertising messages about Evista, finding that "PCPs are more promising . . . indicate they will put more patients on Evista . . . and think the breast cancer data is more compelling."
Although the FDA had told the company it found "critical problems" with the Evista study and didn't approve the drug for preventing breast cancer, according to FDA records, it allowed Lilly to make a minor change in the drug's label. The company inserted three sentences that detailed the precise numbers of breast cancer cases from the study as well as the following: "The effectiveness of (Evista) in reducing the risk of breast cancer has not been established."
That was all Lilly needed to do. Company officials got the word to 1,000 sales representatives. In its plan "Maximizing the Breast Cancer Label Change," it scripted answers to deal with expected concerns, including the "relatively small number of cases" in its study.
The new sales strategy irked one of Lilly's competitors, now known as AstraZeneca PLC, the maker of tamoxifen, which the FDA has approved to prevent breast cancer.
The British company did its own survey of primary-care and women's-health doctors and found that nearly 65 percent of them learned about Evista from sales reps, not the medical literature. Another survey showed that 22 percent thought Evista was FDA-approved for preventing breast cancer. Many already were prescribing it that way.
AstraZeneca took the case to court, demanding that Lilly stop such practices and divulge the "call notes" its sales representatives wrote after visits to doctors. The case was settled, and the details are confidential. Evista someday may prove to be good for preventing breast cancer, although Lilly said it was at least two years away from completing the study that may allow it to apply to the FDA for such a use.
As for those call notes, some were illuminating.
"Told MD about the new indication of EV to prevent breast cancer," one sales rep wrote in December 1998.
One physician told the sales rep he was "extremely excited about the breast data!! Said that if his wife were eligible, he would put her on Evista."
"Sit down detail w/MD," another note read. "Went through entire Evista message w/disclaimer at first to let him know I could not talk about stuff that was off label unless he asked questions first. It worked. He asked all about breast cancer."
(researcher Tish Wells contributed to this report.)