Why do clinical trials exclude depressed people?

A man looks at the Mediterranean Sea as pigeons fly past him on a foggy morning in Tel Aviv November 16, 2010.

NEW YORK (Reuters Health) - When the U.S. Food and Drug Administration announced in 2009 that Pfizer Inc's smoking-cessation drug Chantix would need to carry a restrictive "black box" warning label, the move didn't really surprise the market.

The drug's sales had already been declining. By the end of 2009, they had dropped to $700 million, down from $846 million the previous year.

"When this drug launched, a lot of people expected a blockbuster drug, with over $1 billion in sales, but then the reports of the side effects started coming in," Damien Conover, an analyst with Morningstar, told Reuters Health.

The FDA's response came after hundreds of reports of erratic behavior and several suicides. Now, Pfizer faces a civil lawsuit involving at least 1,200 patient complaints.

In retrospect, experts say all of this could have been avoided -- or at least predicted -- had the company's clinical trials been designed differently. Pfizer tested Chantix on thousands of smokers in order to get FDA approval, but the clinical trials excluded people with depression.

That's despite the fact that the disorder is especially common among smokers: More than 40 percent of smokers are depressed, compared to just 7 percent of the general population, according to statistics from the Centers for Disease Control and Prevention and the National Institute of Mental Health.

Doctors have long known that depression often coincides with other disorders, such as addiction and heart disease. Yet clinical trials for these related conditions routinely exclude patients with depression.

This strategy may benefit drug companies during the early testing stages, but it can backfire once the treatment is released into the real world.

"Adherence is often an issue with depressed participants," said Dr. Anne Thorndike, an internist who studies smoking cessation at Massachusetts General Hospital in Boston. "But how generalizable can the results be if depression is excluded?"

COMMON PROBLEM

Pfizer's decision to exclude smokers with depression was not unusual, and was within FDA guidelines.

Sanofi-Aventis did the same thing with the weight loss drug rimonabant (Acomplia) and saw the drug banned in Europe after being linked to suicide shortly after it was approved there a few years ago.

Last summer, an FDA panel rejected Boehringer Ingelheim's "female Viagra" drug, flibanserin, in part because it wasn't tested on women with depression. The company discontinued work on the drug a few months later.

A recent search by Reuters Health of the federally run ClinicalTrials.gov database revealed the practice is common in smoking cessation studies.

Of 38 actively recruiting large-scale late-stage studies known as Phase 3 clinical trials, 21 excluded people with mental illness, and 10 did so for depression specifically.

Late-stage studies of heart disease were the same: 27 of 154 studies, or about 17 percent, excluded patients with mental illness.

In some trials, the exclusionary language was quite specific. For example, a heart disease study in Cleveland only excluded participants taking certain kinds of antidepressants.

In others, the exclusion criteria were vague. Several trials excluded participants with "significant" psychiatric disorders, and a few avoided the topic altogether, with disclaimers warning that anyone deemed incapable of adhering to the study protocol could be excluded.

Clinical trials are supposed to mirror the potential patient population as much as possible, but drug companies frequently screen out participants with a history of psychiatric illness, citing safety concerns and fears that their disorder will interfere with the study protocol or cloud the results.

"For drug companies, there's always a balance between trying to limit the study population to those people who will have minimal side effects and have the best treatment response," says Dr. Frank Greenway, an obesity specialist who conducts clinical trials at Pennington Biomedical Research Center in Baton Rouge, Louisiana.

Researchers have used similar reasons in the past to justify the exclusion of women and the elderly from clinical trials.

Excluding depressed patients may have negligible effects on some studies. For example, a trial of a treatment for a rare blood disorder. But given how common depression is, some researchers argue such exclusions make little sense in studies of diseases that affect the general population, particularly those that are known to overlap with depression, such as addiction, obesity, chronic pain, and heart disease.

Excluding patients who are more likely to have serious side effects will probably help limit the number who drop out before the end, wrote the authors of a recent article on clinical trial guidelines in the journal Pain.

"Nevertheless," they said, "the greater the number of exclusion criteria in a clinical trial, the less likely its results will be generalizable to the population of patients in the community from which its sample is drawn."

UNINTENDED CONSEQUENCES?

The exclusion of depressed participants could be inadvertent in some cases.

"Clinical trial design is time-consuming, so there's a temptation to copy and paste exclusion criteria from one study to another," said Brian Egleston, a biostatistician at the Fox Chase Cancer Center in Philadelphia, Pennsylvania, who has studied the exclusion of gays and lesbians from clinical trials.

For example, a cancer researcher studying smoking cessation might recycle exclusion criteria from a cancer study and wind up excluding patients with psychiatric disorders without meaning to. "You sort of hope researchers will be thorough and thoughtful, but that doesn't always happen," he said.

To Dr. Gregory Simon, an obesity researcher at Group Health Research Institute in Seattle, Washington, excluding participants with depression from weight loss studies amounts to discrimination because it denies them access to potentially helpful treatments.

"People with depression are very often excluded from studies of treatments for obesity, and people with bipolar disorder are almost universally excluded, but there's no reason these patients shouldn't be studied," he said.

Other researchers note that depression can have an effect on treatment response. For example, a study that Thorndike at Mass General published in 2008 in the Archives of Internal Medicine showed that smokers with depression were 22 percent more likely to relapse.

And last fall, researchers at the U.S. Department of Veterans Affairs published a study in the Archives of Surgery linking depression to increased complications during surgery. There is also evidence that certain drugs, including beta-blockers and some sleep medications, may worsen depression symptoms.

FUTURE OF CHANTIX

Pfizer will soon have a better idea of how depression and other psychiatric disorders affect the way people respond to Chantix. In addition to the black box label, the FDA is requiring the company to conduct a large follow-up study on participants with mental health problems.

The company is complying, but representatives have also argued in the media that nicotine withdrawal could be responsible for the psychological effects associated with the drug. "No causal relationship has been established," Pfizer spokesman MacKay Jimeson said. Jimeson confirmed that the company is starting the new Chantix study, but it's unclear how long the trial will take to complete.

Thorndike predicts that eventually all smoking cessation trials will have to include patients with depression, for practical reasons if nothing else. The FDA is also in the process of putting together a guideline for assessing suicide risk during clinical trials -- the first draft was published in September -- that could help alleviate safety concerns over including participants with psychiatric disorders.

"So many patients are on antidepressants now. Researchers can't be too picky," said Thorndike.

(Editing by Ivan Oransky)