On Thursday night, Pfizer Inc. Chief Executive Henry McKinnell got an unexpected phone call at his home in Greenwich, Conn. He learned a review of a lone cancer study had for the first time linked high doses of the companies Celebrex painkiller to greater heart-attack risks.
Similar concerns had caused Merck & Co. to pull its blockbuster painkiller Vioxx from the market at the end of September. But after conferring with his colleagues, Mr. McKinnell quickly concluded: Pfizer would keep selling Celebrex.
By keeping the drug on the market for now, Pfizer is betting that the medical community and consumers will ultimately decide theres still a valid need for a Cox-2 inhibitor, the class of painkillers that includes Celebrex. Its a huge gamble. If new adverse information comes to light, it could subject Pfizer to an intense legal attack. Its far from clear whether doctors will continue to prescribe the drug, which this year is expected to have more than $3 billion in sales. And the Food and Drug Administration has indicated it will take a hard look at the entire class of drugs.
We do have great concerns about this product and this class of products, said Acting FDA Commissioner Lester Crawford on Friday. He said all regulatory options including withdrawal are open, but that the agency hasn't reached a decision on Celebrex.
The case goes to the heart of a major dilemma facing pharmaceutical giants and the government officials who regulate them. In sufficient doses, nearly any drug can cause harm to some subset of patients. The trick is to weigh whether that risk offsets the drugs potential benefits.
The most important thing is that we don't allow the confusion and uncertainty to push us to a premature decision, said Mr. McKinnell in an interview. These are complicated medical and scientific issues.
In the case of Celebrex, Pfizer argues that any heart problems are uncommon and the dosage people tend to use is much lower than that given in the cancer study. A second cancer study evaluated by the same group of cardiologists found no link between Celebrex and heart risks. And doctors and regulators have to weigh the risks of alternatives, such as potentially fatal stomach bleeding, with some other widely used painkillers like aspirin and ibuprofen. Indeed, the FDA has pointed out that the cardiovascular profiles of these nonsteroidal anti-inflammatory drugs, or NSAIDS, haven't been studied with the same rigor as the Cox-2 pills, so there may be unknown danger for the older drugs.
Pfizers announcement on Friday of the findings sent shareholders, regulators, and patients and their doctors scrambling to figure out the medical and corporate implications.
Investors, concerned over a potential hit to Pfizers earnings and future liability over lawsuits, pummeled its shares, which fell 11%, or $3.23, to $25.75 on the New York Stock Exchange. Nervous doctors looked for alternatives. Plaintiffs attorneys, some of whom have already filed Celebrex lawsuits, felt invigorated by the latest study and said they would be more aggressive about their Pfizer cases now.
Were going to be pushing much harder on our Celebrex cases, said Andy Birchfield, an attorney at Beasley Allen Law Firm in Montgomery, Ala., which had filed 53 cases against Pfizer over Celebrex more than a year ago. Mr. Birchfield said the firm is evaluating an additional 325 potential cases.
Still, some lawyers say the case against Pfizer appears to be weaker than cases against Merck over Vioxx, in part because the steady drumbeat of problematic studies over Vioxx in the past few years well outnumbered those about Celebrex.
Just a few months ago, Pfizer was boldly crowing that Celebrex was free of the cardiovascular risks seen with Vioxx. But the cardiologists analysis linked Celebrex taken at high doses to a greater than 2.5 times risk of heart attacks, strokes or death from cardiovascular causes compared to a placebo pill. That level of risk appeared even greater than the one found in patients taking Vioxx in a similar trial that led Merck to withdraw Vioxx. That study, called Approve, detected a doubling of risk of heart attacks and strokes in patients taking the pill daily for longer than 18 months.
In an unusual move that reflects serious concerns about the new results, the FDA asked Pfizer to suspend its use of direct-to-consumer advertising and alter its marketing to doctors, as regulators and the company examine the data from the National Institutes of Health trial.
It was a desire not to have mixed messages going out to physicians and patients, said John Jenkins, director of the FDAs office of new drugs. We thought it would be very strange for consumers to be watching the evening news and see a story about Celebrexs potential risk and then see an ad with a contrasting message.
A Pfizer spokesman said the company has agreed with the FDA to immediately suspend television, newspaper and radio advertising of Celebrex to consumers. Some magazine ads, already in the works, may still appear due to production lead times. The company also agreed to modify the messages on Celebrex that its sales representatives communicate to doctors, as the FDA requested.
During a 10 p.m. conference call last Thursday of Pfizer executives, medical personnel, and lawyers, Mr. McKinnell posed the question raised by Mercks withdrawal of Vioxx. I asked if anyone on the call believed should the drug be withdrawn and nobody did. Mr. McKinnell agreed.
For starters, he said, Celebrex is a very important agent for patients who are dealing with pain or arthritis, and this is either the best option for many of them or the only option. The National Cancer Institute also thinks the drug at extremely high doses could be useful for cancer patients.
As for the precedent set by Merck, Mr. McKinnell said: Vioxx made us alert to this risk. We had early signals of cardiovascular risk with Vioxx. We saw none of that in our data for Celebrex. This new information was inconsistent with the great body of information wed accumulated over 10 years. And it was contradicted by a very large study analyzed by the same group of specialists. But we still haven't received the data; Its somebodies summary of the data. Only a handful of the people in the world have seen the full data.
Mr. McKinnell acknowledged the company considered the issue of legal liability during deliberations late Thursday night. The legal discussion centered on requirements of the FDA and securities regulators, he said. The best way to avoid liability is to do the right thing. In todays world you cant avoid lawsuits but you can avoid liability, he said.
Pfizer has a lot riding on the success of the Cox-2 franchise. The drug maker faces generic competition to a slew of blockbusters, including antibiotic Zithromax and antidepressant Zoloft, with current sales of $14 billion, by the year 2007. Pfizer bought Pharmacia Corp. last year for nearly $60 billion because of Celebrex, which has quickly become one of Pfizers biggest sellers, with $2.3 billion in revenue through the third quarter. Another Pfizer Cox-2, Bextra, had sales of nearly $1 billion for the same period. Bextra recently has also been linked to heart problems.
There have long been questions about whether the Cox-2 drugs were all tied to a risk of cardiovascular problems like heart attack and stroke, but the evidence has been mixed. Some scientists, most prominently Garret FitzGerald from the University of Pennsylvania, have argued that all the Cox-2 drugs, including Celebrex, could be expected to cause cardiovascular problems by making blood more prone to clotting. Blood clots can lead to heart attacks or strokes.
Earlier research suggested that Celebrex may not have shared Vioxxs cardiovascular risks. At the same time that Merck did the study known as Vigor, which led to worries about Vioxxs cardiovascular risks, there was a similar trial done with Celebrex. Though there were so
me key differences in the designs of the two trials, both were large studies intended to show benefits for patients gastrointestinal health tied to the Cox-2 drugs, when compared with older painkillers.
An FDA reviewer who examined the Celebrex trial, known as Class, concluded that the overall cardiac adverse event profile for Celebrex was not clearly different from that of the older painkillers diclofenac and ibuprofen. The data didn't support a large adverse effect of Celebrex in causing serious cardiovascular problem. But, he cautioned, they don't exclude the possibility that Celebrex promotes blood clotting in a less apparent way. It would take larger studies to find out if this existed.
But in 1999, the National Cancer Institute launched a study that involved patients taking high doses of Celebrex400 milligrams (200 milligrams twice a day) and 800 milligrams (400 milligrams twice a day)to see if the drug could prevent precancerous growths called polyps from progressing to colon cancer. The average duration of treatment with Celebrex or placebo was 33 months.
Scientists had known that aspirin and related drugs can help prevent this progression. But as with pain relief, the drugs can cause bleeding, particularly at high doses. The NCI wanted to test whether Celebrex could protect against cancer without undue bleeding. At the time, people weren't concerned about heart problems, an issue that hadn't surfaced yet.
We didn't design the study to be a detailed analysis of cardiovascular endpoints, said Monica Bertagnolli, the trials principal investigator and also associate professor of surgery at Harvard Medical School.
Thats one reason why the link to heart problems was almost missed. After Vioxx was linked to higher heart risks, Dr. Bertagnolli asked the studies official safety panel to evaluate her studies data for any Celebrex heart problems. The panel of cancer experts gave it a clean assessment. Still uncomfortable, Dr. Bertagnolli took it one step further and called on Marc A. Pfeffer, a cardiologist and Harvard colleague, for advice.
Dr. Pfeffer has long experience at Brigham and Womens Hospital, where Dr. Bertagnolli is a surgeon, running clinical trials for cardiovascular drugs as well as evaluating and confirming side effects and serious events such as heart attacks, a sometimes-tricky determination even among well-trained experts.
After their conversation, and after consultations with other trial leaders and the NCI, Dr. Pfeffer was asked to head a small panel of cardiologists to review safety data for the trial.
Meantime, Pfizers own cancer-prevention study, begun by Pharmacia in early 2001 before the companies merged, was designed to see whether Celebrex could prevent the recurrence of precancerous polyps in patients who had already had growths surgically removed. The trial included more than 1,500 patients who were randomly assigned to take either a placebo or 400 milligrams of Celebrex once a day. When Pfizer learned of plans to investigate the potential for heart risk in the NCI study in early October, it asked that its own study be evaluated by the same team.
The group, which called itself the cardiovascular safety committee, set up a protocol to look for deaths from cardiovascular disease, heart attacks and strokes, essentially the same group of problems that led to the withdrawal of Vioxx. Scott Solomon, another cardiologist at Brigham and Womens hospital, led the effort to track down hundreds of medical records at 90 participating medical centers in four countries the U.S., Canada, Great Britain and Australia and, without knowing whether patients got Celebrex or placebo, determining which ones suffered a relevant adverse event.
The six-week effort turned up 41 patients in the NCI study who fit the criteria and 27 in the Pfizer study. The question was, how many of the patients who suffered the adverse event were taking Celebrex and how many the placebo.
The cardiologists summoned the independent safety monitoring boards for each of the trials to a meeting at a hotel adjacent to New Yorks La Guardia Airport on Dec. 10, where the results were first presented.
For the NCI study, it looked very much like the Vioxx story, Dr. Pfeffer said. In our perspective, we felt that the [risk] was very comparable to that seen with Vioxx. The level of risk increased with dosage.
But there wasn't any increased risk for Celebrex in the Pfizer data. Given the potential impact of their findings, the cardiologist took a few more days to double-check the results before issuing a final report to the same panels on Thursday.
The findings: in the NCI study, six, or 0.9% of placebo patients had suffered a serious event, compared with 15 patients, or 2.2% of those on 200 milligrams of Celebrex twice a day, and 20 patients, or 3% of those on 400 milligrams twice a day.
In the Pfizer study, 11, or 1.8% of placebo patients either died from cardiovascular disease or suffered a heart attack or stroke compared with 16, or 1.7% of those taking 400 milligrams of Celebrex once a day.
The absolute risk was small, said John McMurray, a cardiologist at University of Glasgow, Scotland, and a member of the special committee, but, speaking to the data from the NCI study, he adds: If you take account of the vast number of patients taking these drug, then the absolute risk, if its real, would translate potentially into a big public health problem. About 28 million people world-wide have taken Celebrex.
The cardiologists delivered the data to safety officials overseeing each of the trials, recommending that the investigators in the NCI study stop giving patients Celebrex.
The NCI agreed. But the patients were due to stop taking the medication soon anyway, and researchers plan to evaluate the data to see whether the drug had any effect on colon cancer.
The finding shocked Pfizer executives. A few hours after darkness fell in New York Thursday, Joseph Feczko, president for drug development, got a call from a doctor on the team at Pfizer that works on Celebrex. Dr. Bertagnolli had called Pfizer oncologist Craig Eagle, who was working on the cancer prevention project, to arrange an emergency conference call with the NCIs Ernie Hawk, among others, to discuss the findings.
Shortly after 7 p.m., when the call with NCI was over, Dr. Feczko called Pfizer Chief Executive Mr. McKinnell at home with the Celebrex news. During a short conference call at 8 p.m., Mr. McKinnell ordered that the companies regulatory team contact the FDA to let them know what was coming. In a second conference call at 10 p.m., Mr. McKinnell doled out assignments, including who would write the companies statement for release Friday morning.
Pfizer has had experience managing side-effect issues before. Viagra, its drug to treat impotence, also was linked to heart problems when combined with drugs to treat angina, and the company managed that curve ball with disclosure of the problem in its labeling, avoiding major fallout from litigation.
A decision to withdraw a drug has to be made in the context of all the data not just one study, said Mr. McKinnell, the companies chief executive. Mr. McKinnell said he takes 100 to 200 milligram doses of Celebrex for acute inflammation and that he would continue to do so.