fb tracking

Letter Revealing Heart Dangers in an Unpublished Clinical Trial of Celecoxib (Celebrex)

Dr. Lester M. Crawford, Acting Commissioner
Food and Drug Administration
5600 Fishers Lane
Rockville, MD 20857

Dear Dr. Crawford,

Since filing our petition last week (January 24th) to immediately ban celecoxib (Celebrex) and valdecoxib (Bextra)[1] we have discovered the results of an unpublished randomized placebo-controlled study by Pfizer, finished more than four years ago, that showed a significantly increased rate (3.6-fold) of serious cardiovascular adverse events and more than a doubling in the rate of cardiovascular deaths in people using celecoxib compared to those using a placebo in a study concerning Alzheimer’s disease.[2] This letter supplements last week’s petition.

Despite these older, but previously undisclosed findings, when Pfizer recently announced the results of an ongoing colon polyp prevention trial that was stopped because of an 2.5 to 3.4-fold increased risk of fatal or non-fatal major cardiovascular events in patients using celecoxib, Pfizer CEO McKinnell stated that “these clinical trial results are new.”[3]

Pfizer Study IQ5-97-02-001

The primary outcome measure of this Pfizer study was to determine if celecoxib would alter the progression of Alzheimer’s disease. Patients were randomized to get either 200 mg of celecoxib twice a day or a placebo and were followed for one year. 285 patients received celecoxib and 140 patients received a placebo. According to Pfizer, “oral doses of celecoxib 200 mg BID [twice a day] for a 52-week period did not statistically significantly limit or attenuate the symptomatic progression of Alzheimer’s Disease.”

The table on page 6 of the unpublished results2 of this study (finished in 1999, results completed in December 2000) is a listing of “Serious Adverse Events Occurring in More than one Patient.” I have extracted from this table all of those serious adverse events concerning the cardiovascular system and this table is shown below.

Even though the study was not large enough (powered enough) to demonstrate a statistically significant increase in any individual serious cardiovascular event, an analysis of a composite of all cardiovascular events, shown below, reveals an increase in those patients getting celecoxib.

The combined rate of all serious cardiovascular adverse events in patients getting a placebo was 2.1% but was greatly increased in those getting celecoxib to 7.7%, a 3.6-fold increase in cardiovascular risk in those people taking celecoxib. (p=0.03)

Serious Adverse Cardiovascular Events Occurring in More Than One Patient

Placebo (n=140)

Celecoxib (n=285)

Cerebrovascular disorder (stroke)



Cardiac failure



Pulmonary edema



Angina pectoris



Atrial fibrillation (cardiac arrhythmia)



Myocardial infarction



Total (rate per patient)

3 (2.1%)

22 (7.7%)

Thus, there was a statistically significant increase in the composite of all serious cardiovascular events in patients getting Celebrex compared to patients getting a placebo.

There were two deaths (out of 140 patients) in which cardiovascular diagnoses were mentioned in the placebo group and nine deaths (out of 285 patients) in which cardiovascular diagnoses were mentioned in the group getting celecoxib. This also represents a more than 2-fold increase in the rate of cardiovascular deaths in people getting celecoxib compared to those getting a placebo (from table on page 7 of the Pfizer results).

When the results of this study were first posted on the PhRMA web site around the beginning of this month, there was no mention of the statistical significance of these cardiovascular adverse events. But when the posting was modified last Monday (the day our petition to ban celecoxib and valdecoxib was filed) the company admitted that “A statistically significant difference favoring placebo in adverse events was observed for certain CV-related body system terms (Cardiovascular Disorders,General; Heart Rate and Rhythm Disorders; Myo, Endo, Pericardial & Valve Disorders).”

Although I have learned that the FDA now has possession of this study, I do not know when Pfizer forwarded the data from this study (IQ5-97-02-001) to the agency. There was certainly no mention of this study in February, 2001 when there was an FDA hearing concerning the results of the CLASS study, finished in 2000, that failed to show any gastrointestinal advantage over older NSAIDs for celecoxib. Nor has FDA ever mentioned it in the context of all of the public discussion of COX-2 dangers, especially in the past four months. If the FDA has had this study for a considerable amount of time and did not make the results public, it has seriously failed, once again, in its mission to protect the public health. If, as is entirely possible, Pfizer did not send these results to the FDA until after the December 17th announcement of similarly increased cardiovascular risk in the polyp-prevention study, the company has set itself up for enormous liability for having failed to inform the outside world of these important findings.

Given that there were 36 million prescriptions written in the U.S. for Celebrex and Bextra last year – thus about three million prescriptions a month – the FDA must act immediately to protect people from these two dangerous drugs by removing them from the market. I also look forward to your response to the question as to when Pfizer forwarded this critically important information to the FDA.  


Sidney M. Wolfe, M.D.
Public Citizen’s Health Research Group


[1] Public Citizen 1/24/05 petition to the FDA to immediately ban celecoxib and valdecoxib. https://www.citizen.org/Page.aspx?pid=2396

[2] PhRMA web site at:    http://www.clinicalstudyresults.org/. This web site (http://www.clinicalstudyresults.org/drugdetails/?inn_name_id=55&sort=c.company_name&page=1&drug_id=76) was first started on October 1, 2004 and the Celebrex study was finished, according to the site, on December 22, 2000.  The site states that there have been “No Publications” from this study. The study was entitled: “Results of a double-blind, randomized, placebo-controlled study of celecoxib in the treatment of progression of Alzheimer’s Disease.” Since data was originally downloaded data from the site, Pfizer has made a change (dated 1/24/05) on the last page of the results now stating that the study was “based on reports completed December 22 2000 and December 22 2004”. When I first observed the site there was no mention of December 22 2004, only December 22 2000 for the completion of the report. As of the 1/24/05 “addition” to the results of the study, Pfizer now admits that “A statistically significant difference favoring placebo [compared to celecoxib] in adverse events was observed for certain CV-related body system terms (Cardiovascular Disorders, General; Heart Rate and Rhythm Disorders; Myo, Endo, Pericardial & Valve Disorders).”

[3] Pfizer statement on December 17, 2004.