Linda Ware with her granddaughter, two days before she committed suicide in 2008.

Late one morning in June 2008, 57-year-old Southern California real estate agent Linda Ware was driving with her cousin along a desert highway when she began hallucinating. Envisioning in the distance a sign that read “God Is in the Realm,” she pulled over suddenly and ordered her cousin out of the car. Then, just as abruptly, Ware burst out laughing and pulled back onto the road again.

Although Ware suffered from depression, as her daughter Cary Ussery related, she’d never acted like this. A few days earlier, however, she had started taking Chantix, a pill meant to help her quit smoking by suppressing the effects of nicotine on the brain.

The day after the driving incident, a family friend found Linda Ware slumped by her bed, dead from a fatal cocktail of prescription drugs, a suicide note at her side.

Tragedy has plagued Chantix ever since it was approved in May, 2006, even as the drug has helped some smokers kick the habit. By mid-2009,  the U.S. Food and Drug Administration had received reports of nearly 100 suicides, 200 attempted suicides and close to 5,000 serious psychiatric events overall. Hundreds of reports of side effects have continued to stream in this year.

A review of the drug’s history shows that Pfizer Inc., the giant pharmaceutical company that makes Chantix, failed in its years of clinical trials to test the product on the mentally ill or those with a recent history of depression — even though millions of smokers suffer from psychiatric problems. Moreover, FDA regulators approved Chantix after a speeded-up “priority review” process, and did not request a follow-up study on mentally ill patients using the medication, even though the agency’s own safety reviewer reported that the exclusion of such smokers may have undermined the clinical trials.

It wasn’t until three years later, after thousands of reports of serious side effects, that the FDA told Pfizer to conduct trials including people with a history of mental illness. The agency then also slapped a so-called black box warning, the FDA’s strongest alert, on the medication.

The controversy over Chantix’s side effects, particularly for smokers with mental health issues, has triggered a torrent of lawsuits. About 1,000 such cases have been filed in federal court, and plaintiffs lawyers say they anticipate bringing forward more than 1,000 additional suits. There also are scattered cases in state courts in New York, Illinois and elsewhere. Plaintiffs in more than half of the cases claim the drug led to suicides or suicide attempts, and many say they suffered psychosis, blackouts, aggression, diabetes or other problems. No cases have yet been tried or settled.

The plaintiffs lawyers argue that Pfizer, the world’s largest pharmaceutical company with sales this year of more than $60 billion, neglected to test Chantix adequately before its release, deliberately hid evidence of serious side effects and failed to sufficiently warn consumers about its risks. Pfizer has turned over to plaintiffs lawyers more than six million pages of documents under a protective order that bars their public release.

Pfizer, for its part, defends Chantix and says it “acted responsibly and appropriately at all times in connection with the development, approval, and marketing” of the drug. The FDA, likewise, says it acted properly in approving the drug, despite the problems that emerged after it went on the market.

“The agency does not feel any mistakes were made,” said FDA spokeswoman Sandy Walsh.“We can never speculate as to what may happen with a drug once it goes into widespread use after approval.”

The FDA’s failure at the outset to require Pfizer to include the mentally ill in its research, however, points to a potentially serious flaw in the agency’s regulation of drug safety for these particularly vulnerable consumers. Although the mentally ill commonly are excluded from drug clinical trials, they account for a disproportionate number of smokers and are a key part of the target market for smoking cessation treatments.

Their exclusion baffles experts such as Dr. Karen Lasser, a Boston University researcher who has studied the link between mental illness and smoking. “You need to think about who is going to be taking the drug,” she said.

Paul Thacker, a former Senate Finance Committee investigator who has written about FDA decision-making, agreed. “If you’re not thinking in that way,” he said, “then you’re not doing your job.”

Back when Chantix was approved, Pfizer officials figured they had a potential major new star in their portfolio. With nearly 70 percent of the estimated 45 million smokers in the U.S. hoping to quit, the market appeared to be vast. Sales quickly zoomed to $883 million in 2007, which turned out to be the drug’s peak year. (During the first nine months of this year, sales totaled $522 million.)

Chantix, which is sold as a pill, works by taking the pleasure out of smoking by interfering with the way nicotine ordinarily affects the brain. At the same time, it spurs the release of dopamine, which helps control the brain’s pleasure centers the way smoking usually does.

Pfizer launched Chantix with a news release spotlighting the dire statistics on quitting smoking: less than 7 percent of smokers who try to quit on their own make it past the one-year mark. In contrast, clinical trials showed that about 22 percent of Chantix users who took the drug for three months were able to abstain from smoking for a year or more.

The company early on promoted its product through a series of TV, print and online ads that, without mentioning Chantix or its side effects, sought to sell people on the idea of quitting smoking. The “My Time to Quit” campaign was intended to draw smokers to a website that, after providing quick facts about kicking the habit, tiptoed to information about the drug. Overall, Pfizer has spent about $300 million on advertising for Chantix, according to the Nielsen Co., a media information business.

Pfizer also spread its influence by paying doctors to give talks to other physicians on smoking cessation techniques while also giving funds to universities. The University of Wisconsin-Madison said the company has given it more than $3 million, mainly for continuing medical education courses it offers on smoking cessation.

The multi-pronged promotional effort by Pfizer eventually addressed rising concerns about the drug. In 2008, Joseph Feczko, then Pfizer’s chief medical officer, wrote an opinion piece in the Wall Street Journal. Headlined “Smoking Has Side Effects Too,” the piece tried to offset the negative publicity about Chantix with a reminder on the hazards of smoking. It also explained how regulators and drug companies conduct follow-up research, and alert the public, after they receive reports of serious side effects.

Along the way, a controversial tack Pfizer took to push its product was a study written by a team of employees headed by Kathryn E. Williams, then part of the company’s Global Research and Development unit. The study, which was published in 2007 in the journal Current Medical Research and Opinion, deemed Chantix safe for long-term use. (Labeling on the drug prescribes a 12-week course of treatment, but the study looked at consumers who used Chantix for a year.)

The article, however, was flawed in that failed to provide statistical analysis to back up the claim, according to Dr. John Spangler, a smoking cessation expert at Wake Forest University’s School of Medicine. He wrote to the journal to complain, noting that the numbers actually indicated Chantix patients were 2.5 times more likely to suffer a serious adverse event than those on a placebo.

“I don’t know what they were trying to accomplish,” Spangler said in a recent interview. “A faulty study has been included into the world’s medical literature.” Pfizer declined to comment on the study.  The journal, for its part, said via email that the article “underwent rigorous and independent peer review, and full disclosure of the authors’ employment and funding was made.”

Other studies have demonstrated that Chantix has helped some smokers, but often is no better than other options for quitting. For instance, a study published in in the journal “Thorax” in 2008 showed that Chantix works better than the nicotine patch for some people, but not for others. Other researchers have concluded that nicotine gum is just as effective as Chantix.

What’s more, most people who give up smoking quit cold turkey, without medication. A 2006 survey of more than 8,000 smokers by the National Cancer Institute found similar success rates among those who used medication in trying to quit and those who didn’t. The figures showed that 16 percent of the non-medicated group abstained from smoking after nine months, versus 14 percent for the group using medications.

Still, experts such as Lirio Covey, a professor of clinical psychology at Columbia University who worked on early studies of Chantix, say the FDA’s approval of Chantix made sense. “Some people really have a hard time stopping smoking. Chantix does have some utility,” Covey said. “I would say that it’s a pretty good drug, but it’s not the first resort” for treatment.

For Covey, who said she has received close to $80,000 from Pfizer this year to study the effects on mood of smoking cessation, the key facts are these: millions have taken Chantix to stop smoking, and the number of people who suffered severe side effects is relatively small. In her estimation, the benefits of the drug outweigh the risks. She also believes it’s still not clear if the side effects are the result of the drug or nicotine withdrawal.

But questions about why Chantix’s safety wasn’t fully evaluated extend back at least to May, 2006, when the FDA’s safety reviewer for the drug, Dr. Howard Josefberg, submitted his report to the agency. He endorsed approving Chantix, but found that Pfizer’s clinical trials may have been “too carefully screened.” He noted that the studies excluded people treated for depression over the previous 12 months as well as “those with histories of panic disorder, psychosis or bipolar disorder.”

“The patient population studied, then, may not represent the true target population should varenicline be approved,” added Josefberg, using the generic name for Chantix.

The report by Josefberg, who declined to be interviewed for this story, indicated that three or four people died out of the roughly 5,000 smokers treated with Chantix in clinical trials. FDA researchers, however, said all of the fatalities appeared to be unrelated to the drug.

Still, one death in particular piqued the interest of scientists. A white male in his 60s who took the drug for close to six months took his own life. “Patient 103510121069 represents the most potentially concerning case,” Josefberg wrote. He “committed suicide by hanging 27-days after completing the 24-week varenicline treatment.” The patient had a history of severe depression but neglected to share his troubled past with those screening patients, according to FDA documents.

More than 80 other Chantix patients in the trials reported serious side effects. In one case, a 46-year-old white female, who stopped taking the drug after a week, arrived at work speaking incoherently, confronting colleagues, and overturning furniture, according to FDA documents. She was hospitalized for acute psychosis. Later reports indicated the patient had some history of psychotic behavior, but had failed to tell investigators, according to FDA documents.

All told, Josefberg concluded that there was no clear connection between Chantix and any deaths or serious adverse events, psychiatric or otherwise. He recommended the FDA approve the drug with a warning that alerted users to side effects such as potential heart problems, nausea, insomnia and abnormal dreams, while adding that the trial data was far from conclusive on the drug’s cardiovascular effects.

The routine exclusion of mentally ill subjects from clinical trials, researchers say, is justified in certain cases. Sometimes the subjects are too sick to consent to participate. Mentally ill patients may also fail to follow the proper dose regimens or other instructions. In addition, including people with mental or other health problems in drug research can complicate evaluating the results of a medication.

Yet many experts argue that excluding the vast numbers of Americans who are mentally ill from the trials leaves in doubt the real world effects of drugs on many people who may be vulnerable. According to the National Institute of Mental Health, 26 percent of Americans 18 and older, or close to 60 million people, suffer from a diagnosable mental disorder in a given year. About 6 percent suffer from a serious mental illness.

Beyond that, the mentally ill account for a big proportion of American smokers. A report this spring from the U.S. Centers for Disease Control and Prevention found that 43 percent of adults with depression smoke, versus 22 percent of other adults.

Boston University’s Lasser, in a recent analysis relying on the broader definition of mental illness, concluded that people with such disorders represent about 40 percent of all smokers.

Raymond Lorenz, the author of a recent paper on Chantix and the mentally ill and a faculty member at Auburn University’s pharmacy school, said more research is needed to gauge the risks that psychiatric patients face when they take the medication. Still, he wrote that excluding the mentally ill “seems to be a glaring oversight” in Pfizer’s trials before the FDA approved Chantix.

The FDA could have, but didn’t, ask Pfizer to conduct additional research — in this case, research including mentally ill smokers — immediately after Chantix was approved. The agency sometimes does that as a compromise, to ensure that medications reach patients without extra years of delays, while keeping an eye on a drug’s side effects after it comes into widespread use.

By late 2007, however, the parade of Chantix warnings and restrictions by federal authorities had begun. That November the FDA issued a warning that some patients taking Chantix reported having suicidal thoughts. Two months later, the agency ordered revised labeling indicating that Chantix’s safety for mentally ill patients was not established. Other federal agencies also took action, including the Federal Aviation Administration, which banned pilots and air traffic controllers from using Chantix.

Brandon Campbell, an IT worker at Duke University Hospital.

Finally, in July, 2009, the FDA came around to fully embracing testing Chantix on the mentally ill. “We are going to require that they study folks that have mental health disorders,” said  Dr. Curtis Rosebraugh, an FDA drug evaluation official, in a telephone conference call media briefing. “There is a disproportionate amount [of them] that smoke and they would be potentially exposed to this drug. We have no idea if that subgroup population is at higher risk or not and so we do want to get some sense of that.”

Chantix users like Brandon Campbell already have a sense of that.

In July 2007, Campbell, a 33-year-old IT technician at Duke University Hospital, gave the drug a try after enrolling in an employee wellness program to quit smoking. Although Campbell, like Ware, had a history of depression, there was no indication as yet that the safety of the drug had not been established in people like him. Soon after finishing the recommended 12-week course of treatment, Campbell recalls, he started experiencing severe symptoms. He felt disconnected from reality, had trouble remembering things, and became emotionally unstable—breaking down over trifles. He also contemplated suicide. He’d been depressed before, sure, but had never felt, as he put it, like driving over a cliff.

The problems persisted. This past summer, Campbell said, he sat in his garage in Durham, North Carolina, with the doors shut tight and the car running, hoping to die. Then the police started beating on the door, and he was rushed to the hospital. Campbell still smokes and says he still wants to quit, but he’s exceedingly wary of trying any more pharmaceuticals. “It’s like playing Russian roulette,” he says.