Researchers called Chantix, the smoking-cessation drug linked to psychotic and suicidal side effects, unsafe for patients Wednesday; just days after federal health officials in October deemed the drug safe.

The researchers concluded that 90 percent of reported suicides related to smoking- cessation drugs since 1998 involved varenicline, the active ingredient in the drug made by Pfizer, Inc., the world's largest drug manufacturer.

Researchers concluded that varenicline-based products caused suicidal behavior or depression eight times more frequently than nicotine-replacement products.

We found that Chantix is associated with more suicidal behavior reports than any other smoking-cessation drug on the U.S. market. The risks simply outweigh the benefits, Curt D. Furberg, public health professor at Wake Forest Baptist, said in a statement. Furberg headed the study, joined by researchers from Harvard Medical School and Johns Hopkins University School of Medicine.

The conclusions, published Wednesday in PLoS ONE contradicts those from the Food and Drug Administration that has monitored Chantix since 2007. The FDA already requires the drug manufacturer to place a black box warning for suicidal behavior.

Pfizer said they stand by Chantix and are currently conducting a large double-blind placebo controlled safety clinical trial of Chantix to assess neuropsychiatric safety in patients with and without psychiatric disorders, and results of this large study are expected in 2017, MacKay Jimeson, Pfizer spokesman, said in an email. The federal drug regulatory agency requested the study, according to a spokeswoman.

FDA found no differences between Chantix and nicotine patches in two studies to examine previously-reported psychiatric side effects from Chantix. The FDA announced their results Oct. 24.

Neither study found a difference in risk of neuropsychiatric hospitalizations between Chantix and nicotine replacement therapy, such as NicoDerm patches, Minoo Shafinouri, a pharmacist in FDA's Division of Drug Information said at an Oct. 24 press conference. However, both studies had a number of study design limitations, including only assessing neuropsychiatric events that resulted in hospitalization, and not having a large enough sample size to detect rare adverse events.

In the PLoS ONE study, researchers analyzed 3,249 case reports of serious injury included in the FDA's Adverse Event Reporting System from 1998 through September 2010 for self-injurious behavior or depression linked to varenicline, bupropion (Zyban), an antidepressant approved for smoking cessation, and nicotine replacement products.

The researchers found varenicline associated with 90 percent of cases with suicidal behavior or depression compared with 7 percent for bupropion, and 3 percent for nicotine replacement products.

While suicidal behavior or depression appear to be prominent side effects of varenicline, they are by no means the only safety issues, Thomas Moore, senior scientist at the Institute for Safe Medication Practices and lead author of the study, said in a statement. Varenicline has been associated with aggression and violence in three studies and carries a warning about this behavior. Its effects on vision, cognition and motor control and other risks have led to its being banned for airline pilots, air traffic controllers, military pilots and missile crews, and restricted for truck drivers.

FDA spokeswoman Karen Riley said in an email that the report was not a new finding and that the FDA had previously reported post-marketing events and changed product labeling.

The PLoS ONE study had several limitations, Riley said. The study lacked narratives of individual cases; analyzed portions of the reports rather than observing the frequency of adverse events with use of the drugs, and calculated association rather than measure risk, Riley said.

We continue to believe that when used as directed in the currently approved labeling, Chantix is a safe and effective treatment to help patients stop smoking, Riley wrote.

Pfizer said that Wednesday's announcement amounted to faulty science.

These same authors persist in publishing analyses based on a review of spontaneous reports, Jimeson said. It is important to remember that post-marketing reports do not establish a cause and effect relationship between a medicine and a reported adverse event.  Also, it is not appropriate to draw conclusions based on comparisons between different drugs and reporting rates.  Post-marketing reports can come from any source ranging from patients to healthcare providers, and from phone calls to internet postings and lawyers. Often these reports lack sufficient medical information to enable meaningful assessment.  Because of the many limitations of post-marketing reports, the conclusions made by the authors are not supported by the data.

The authors also reported in July that varenicline was associated with increased risk for serious cardiovascular disease.

The health benefits of quitting smoking are immediate and substantial, Jimeson said. Given the significant public health risks of smoking, Pfizer stands by Chantix as an important treatment option for adult smokers who want to quit.