On September 30, 2004, Merck announced its decision to withdraw Vioxx. In its report, Merck cited the then-new results of the APPROVe study as the reasoning behind its decision. Merck’s analysis of the APPROVe study showed an increased risk of cardiovascular events in patients who had been taking the drug for 18 months or longer. Merck explained that the results of APPROVe were not complete until this time, and that early results of the study had not demonstrated the same cardiovascular risk that was evidenced by the study results as a whole.

In this statement, Merck also said it would conduct further investigations into its other COX-2 inhibitor, Arcoxia, to determine whether that drug’s prescription information should be amended to include further warnings.

Merck also announced that the value of its stock was expected to fall as a result of the news about Vioxx. In the year preceding the Vioxx withdrawal, Vioxx sales had accounted for just over 10% of Merck’s net earnings.

The Martin Report

In 2005, Merck hired the firm of Debevoise & Plimpton to conduct an investigation of all Vioxx study results and communications by Merck, including internal communications related to Vioxx. The firm was paid $21 million for the report, which was called the Martin Report after Judge John S. Martin, Jr., who headed up the investigation. The report concluded that Merck’s senior management had acted in good faith, and that the source of all the misunderstandings about the clinical safety of Vioxx was the sales team’s overzealousness.

The nearly 200-page Martin Report gave a detailed timeline of events surrounding Vioxx and explained Merck’s honest intentions at every turn. The report specifically addressed 27 different allegations against Merck, including its mishandling of APPROVe and other clinical trial results, its aggressive marketing tactics, and the idea that Merck purposefully withheld information about the risks of Vioxx. In each case, the report found that Merck scientists and officials had handled the data in good faith, and that any errors were the result of oversight, not intentional omission of data. The Martin Report did conclude, however, that Merck’s marketing team had exaggerated the safety of Vioxx and had intentionally used sales tactics in place of thorough information.

Read the entire Martin Report on Merck.com. (The report is a pdf so Adobe Reader is necessary to read.)

Diffusion of Responsibility

In response to the Martin Report, which was published in February of 2006, Merck stated that it was reassured that the findings of the report revealed no negligence on the part of Merck scientists or executives. Merck promised to consider the recommendations contained in the Martin Report, presumably by reconsidering and better supervising its advertising strategies. Whether or not Merck ultimately changed its marketing techniques is unclear, but the total spending for the advertising of drugs in general has not decreased.

Merck had hoped the Martin Report would serve as a definitive explanation of Merck’s position on Vioxx, both in the press and in the courtroom. However, in 2006, Merck announced a correction to its original statistical findings in the APPROVe trial.

Merck scientists reportedly made an error in calculating the cardiovascular risk in relation to the passage of time. In light of the corrected calculations, Merck retracted its claim that the risks of Vioxx were negligible during the first 18 months of treatment. This statement blew a gaping hole in Merck’s legal defense, which supported the 18-month benchmark theory, while also hurting the theory that marketing was entirely to blame for the Vioxx problem.

Throughout the investigations of Merck’s role in the handling of Vioxx, lawsuits have swept through courts across the US, costing Merck billions in settlements. The 2006 announcement only served to invite more lawsuits by destroying the 18-month theory and opening up claims of complications from short-term Vioxx users.

Vioxx was released just months behind Pfizer’s competing drug, Celebrex, and both drugs were aggressively marketed to doctors and to consumers. Direct-to-consumer drug advertising is any form of advertising that targets potential users of the drug, such as television, radio, and magazine ads. Merck is reported to have spent more than $500 million on advertising of Vioxx, including maintaining and training a massive sales force.

A WebMD investigation concluded that direct-to-consumer advertising, aimed at inducing consumers to visit their doctors and potentially inquire about advertised drugs, has been shown to cause up to a 2% increase in doctor visits. Additionally, direct-to-consumer advertising attempts to create a positive impression of a new drug in patients to whom it has not yet been prescribed, which increases the likelihood that they will continue taking the drug after the initial trial period.

Examples of Vioxx Direct-to-Consumer Advertising

Vioxx was touted in every form of media as a safe and effective pain killer. Television ads featured figure skater Dorothy Hamill skating to the Young Rascals’ tune “It’s a Beautiful Morning.” Dorothy Hamill appeared in print advertisements as well, endorsing the benefits of Vioxx for osteoarthritis relief. Presumably, if this drug provided relief for someone who pushed the limits of her joints to the degree of a professional figure skater, Vioxx could easily handle the average case of sore joints. Celebrity endorsements are an age-old advertising method, creating unparalleled product recognition. Merck not only used Dorothy Hamill as a celebrity endorser, but the company pursued every other avenue of marketing for Vioxx, from television to radio to online sources.

Free samples are another type of advertising that is common to all drug companies. Free samples allow doctors to introduce patients to new drugs at no cost to the patient, which increases the chance that the patient will continue taking the drug once the free samples run out.

Selling Doctors on Vioxx

While direct-to-consumer advertising is the only kind we see as consumers, doctors are exposed to the most effective marketing techniques of all. Not only do doctors receive free samples, as mentioned above, but they often receive massive amounts of promotional material — pens, tissue boxes, notepads, coffee mugs — which increase name recognition and keep the medications on doctors’ minds. In addition, doctors are visited by pharmaceutical representatives, otherwise known as salespeople. These representatives make up the drug producer’s sales force; often, they are only moderately informed of a drug’s clinical history. Their job is to sell pharmaceuticals, and that starts with getting doctors to write the prescriptions.

According to a critical report in the New England Journal of Medicine (NJEM), Merck’s sales team was trained in subliminal techniques for gaining the trust of doctors and creating the image of a relationship between the doctor and the pharmaceutical company, which in turn made doctors more likely to prescribe their drugs to patients. An example of a subliminal technique is the mirroring of a person’s physical movements, which supposedly increases trust. The NEJM article reminds us that providing thorough and objective information about new drugs should take precedence over making sales.

Such techniques were publicly criticized after the Vioxx withdrawal brought Merck into the limelight. Even the Merck-sponsored Martin Report revealed that Merck’s marketing of Vioxx was overzealous, and ultimately contributed to the drug’s being over-prescribed.

As recently as May of 2008, Merck agreed to pay $58 million to settle a multi-state lawsuit accusing the company of deceptive advertising of Vioxx. If Merck hadn’t yet reconsidered its advertising strategies, this latest settlement surely made the drug giant think again.

In April 2005, the Center for Drug Evaluation and Research (CDER), a branch of the FDA, sent out letters to the manufacturers of all NSAIDs, even over-the-counter (OTC) NSAIDs, urging them to include with their products more detailed information about the cardiovascular and gastrointestinal risks of the entire class of NSAIDs. The statement was directly aimed at the manufacturers of Celebrex, Bextra, and OTC NSAIDs.

This statement was not intended as a call for consumers to stop taking COX-2 inhibitors, but rather for drug manufacturers to assure consumers that their products were well labeled and that patients were provided with accurate information about their risks. It also recognized the need for further research on the safety of selective COX-2 inhibitors.

Statements regarding Bextra and Celebrex

At the same time as it announced these new packaging requirements, the FDA recommended the withdrawal of Vioxx’s sister drug Bextra, produced by Pfizer. As reasons for its decision, the FDA cited safety concerns about Bextra, including unpredictable and potentially fatal skin reactions, and known cardiovascular complications in bypass patients. In addition to bearing these serious risk factors, the FDA noted, Bextra had not demonstrated any advantages over other COX-2 inhibitors. At that time, Pfizer agreed to suspend advertising of Bextra in the US pending a follow-up with the FDA. The drug was withdrawn soon after.

Pfizer’s Celebrex was allowed to retain its FDA approval, leaving it the only available COX-2 inhibitor. Pfizer was instructed to include additional labeling warnings in the Celebrex packaging regarding cardiovascular risks. It was also recommended that Celebrex be used in the lowest effective dosages, and that a clinical trial measuring its long-term risks be conducted immediately.

Searching for the Culprit in the Vioxx Snafu

When Vioxx was withdrawn, Merck was the immediate target of public outrage. All Merck-sponsored clinical studies of Vioxx were examined and re-examined; in several cases, it was discovered that Merck had improperly reported or calculated the cardiovascular events in clinical studies.

Merck was questioned not only for its handling of Vioxx clinical trials, but also for its mass marketing of Vioxx as a safe alternative to aspirin. Merck spent more than $146 million to advertise Vioxx in 2000 alone. The company was criticized for neglecting to mention the cardiovascular risks of Vioxx in its marketing. Many critics claimed that, as a result of massive advertising directed both at consumers and at doctors, Vioxx was grossly over-prescribed. This led some of the blame to be placed on doctors, as well. According to the New England Journal of Medicine, 100 million patients were prescribed Vioxx between its release in 1999 and its withdrawal in 2004.

Whether the fault lay with Merck’s scientists, Merck’s marketing, the FDA, or the prescribing doctors, is hard to determine. It was really a combination of all of these factors that helped Vioxx reach so many uninformed or ill-informed patients.

New England Journal of Medicine Report on Vioxx

In June, 2005, the New England Journal of Medicine (NEJM) published an editorial examining Merck’s marketing of Vioxx. The article criticized Merck for having published incomplete results of the VIGOR study in the publication.

The editorial recounts a meeting between the Arthritis Drugs Advisory Committee of the FDA and Merck in 2001 to discuss the results of the VIGOR study, which had shown an increased risk of myocardial Infarction (heart attack) in patients taking Vioxx compared to those taking naproxen. At this meeting, the FDA argued that this imbalance in the VIGOR results was evidence of the protective anti-platelet effects of naproxen. The FDA concluded that this explanation was insufficient and ordered Merck to notify physicians about the results of the study. In reaction to this meeting, the NEJM article reports, Merck instructed its sales force not to discuss the results of the VIGOR study except to say that VIGOR showed a reduction in gastrointestinal complications.

To comply with the FDA’s orders, Merck created the “Cardiovascular Card,” a short pamphlet that included a non-specific summary of some Vioxx pre-approval trials. The Cardiovascular Card did not mention the VIGOR study and did not include data about cardiovascular events during the trials. The NEJM article claims that the information on the Cardiovascular Card, which was written by Merck’s marketing team, was misleading and only sought to further minimize the apparent cardiovascular risks of Vioxx that had been evidenced by multiple clinical trials.

Further, the article reports that Merck offered special training to its sales team, including using subliminal techniques and other misleading sales tactics, to influence doctors’ overall impressions of Merck. All of this was designed to sell more drugs and to downplay their possible risks. Merck, which had had a solid reputation with the FDA prior to the Vioxx dispute, was now accused of transparent salesmanship at the expense of objective drug information.

The article closes by observing that the questionable marketing practices employed by Merck were most likely representative of drug companies as a whole, and advising that the FDA take a larger role in overseeing the approval process for new drugs.

Off-Drug Follow-ups of Vioxx Trials

In 2006, Merck released results of its off-drug follow-up of the APPROVe study. The research showed that, in the year following treatment, Vioxx patients were 74% more likely to experience a cardiovascular event. Merck claimed that these results were statistically insignificant, but many experts in the medical field disagreed. Merck lawyers still insisted that the risks of Vioxx had only been conclusively demonstrated after the drug had been taken for 18 months.

Cardiologist Steve Nissen of the Cleveland Clinic was among medical professionals who believed these off-drug results were significant, in that they may have indicated that Vioxx caused some kind of arterial damage that might be permanent.

In April of 2008, the Journal of the American Medical Association, or JAMA, reported on the off-drug follow-up results of a Merck-sponsored study called Protocol 078. This study, published in 2004, measured the effectiveness of Vioxx in preventing the symptoms of Alzheimer’s Disease. During Protocol 078, 39 deaths occurred, 24 of which were in the Vioxx group. In the off-drug period, 22 additional deaths occurred, 17 of which were in the Vioxx group. In another study, called Protocol 091, 11 deaths were recorded as non-drug-related, 9 of which were in the Vioxx group. In the opinion of the AMA, these results showed a disproportionate number of deaths among Vioxx users, both during the trials and after they stopped taking the medication.

Merck’s official conclusion on these trials upon publication was that Vioxx was “well tolerated” by patients. Nissen and others in the medical community, as well as representatives of the AMA, disagreed.

Read more about how Merck reacted to the Vioxx withdrawal.

As discussed in Vioxx Clinical History, the worldwide withdrawal of Vioxx was a direct result of the APPROVe study. The APPROVe study was designed to measure the effectiveness of Vioxx in preventing the recurrence of colorectal polyps in post-surgical patients. Before the study could be completed, an unexpected result took precedence…the patients in the Vioxx group started experiencing a disproportionate number of cardiovascular events. Researchers stopped the trial at the same time Vioxx was being withdrawn from the market worldwide.

Merck concluded, in its official report to the federal Food and Drug Administration (FDA) on APPROVe, that the cardiovascular risks of Vioxx appeared after 18 months of treatment. Thousands of lawsuits followed, a majority of which involved patients who had used Vioxx for extended periods. Merck insisted that its clinical trials of Vioxx showed no risk of heart attack or stroke until after patients had taken the drug for 18 months, and this claim was the basis of Merck’s legal defense.

Merck maintained the “18-month benchmark” theory until 2006, when results of the APPROVe study and several other studies demonstrated that there was an equal risk before and after the benchmark of 18 months. Merck admitted in a statement to the FDA that it had misinterpreted the results of the APPROVe study, and that the risk of cardiovascular events with Vioxx appeared as early as 2-4 weeks into treatment. Additionally, the risks appeared to have steadily increased in the Vioxx group, while remaining constant in the placebo group, demonstrating an increasing risk proportionate to the duration of treatment.

There is no question that Merck understood the impact of this statement for its future legal defense. It could no longer maintain the position that short-term use of Vioxx posed no proven cardiovascular risk.

FDA Statements about Vioxx in 2004

When a drug is withdrawn or recalled for safety reasons, the public turns a suspicious eye not only toward the manufacturer of the drug, but toward the FDA, the federal government organization responsible for telling us which drugs are safe and which are dangerous. When the FDA delivered the news of the Vioxx withdrawal and the results of the clinical trials became public, the FDA faced a barrage of angry consumers who wanted an explanation of why they had approved the drug in the first place.

On September 30, 2004, the FDA released its initial statement regarding the Vioxx withdrawal, informing the public that Merck had voluntarily instituted an immediate withdrawal of Vioxx from the worldwide market. The FDA had been informed of this decision during a meeting with Merck officials two days ealier. According to the FDA report, the withdrawal was a recommendation of the DSMB (Data Safety Monitoring Board) overseeing the APPROVe study. The FDA added that, although the overall risk of heart attack was statistically small, Vioxx patients in clinical trials had consistently shown approximately double the cardiovascular risk. Additionally, it was stated that the cardiovascular risk appeared after 18 months of treatment, which explained why the APPROVe trial had not been stopped sooner.

The FDA reiterated that the motivation behind the COX-2 inhibitors introduction was to offer pain relief without the gastrointestinal side effects of traditional NSAIDs. Vioxx, according to the FDA statement, was the only NSAID to have demonstrated the intended gastrointestinally-protective effect.

Next, the FDA briefly reviewed the results of the VIGOR study, which in 2002 had prompted the FDA to mandate a black box warning of cardiovascular events to be added to Vioxx packaging. The DSMB of the APPROVe study concluded that that warning was not enough, and that the cardiovascular risks of Vioxx warranted a worldwide withdrawal of the drug. The FDA reported that it had been in the process of reviewing Vioxx studies in detail when the withdrawal was announced by Merck.

What Happens When a Drug is Withdrawn

In 2004, the FDA published an online Q&A for consumers explaining the background and ramifications of the Vioxx withdrawal. A withdrawal is different from a recall. In a drug recall, patients are encouraged to return any unused portion of their medication to the pharmacy, and pharmacies are instructed to return the drugs to the manufacturer. In the case of Vioxx, a prescription-only drug, any prescriptions that were not filled before the date of the withdrawal (September 30, 2004) could not be filled by any pharmacy worldwide. This meant that patients could choose whether to continue taking the medication until their current prescriptions ran out.

Patients who were currently taking Vioxx were told to contact their physicians before stopping the drug suddenly. Many Vioxx patients switched to one of Pfizer’s sister drugs: Celebrex, which is still available, or Bextra, which was withdrawn shortly after Vioxx, in 2005. Patients who experienced significant pain relief from Vioxx were devastated by its loss, and there was no assurance that Celebrex or Bextra could offer safer alternatives.

From the moment that Merck announced its withdrawal of Vioxx, the company has been subject to countless investigations attempting to unravel the series of events contributing to the Vioxx disaster. Merck’s marketing strategies have been a topic of heated debate, along with the partial or incomplete reporting of several clinical trials.

Merck’s handling of clinical trial results revealing the cardiovascular risks of Vioxx has been the basis of many successful lawsuits. Plaintiffs sought to prove that Merck was aware of the elevated cardiovascular risks of Vioxx before the APPROVe study, and they gained fuel for the lawsuits by looking back at earlier clinical trials that demonstrated the same risks. Not only was there clinical evidence of a 4- to 5-fold increase of cardiovascular events in Vioxx patients as early as 2000, but suits claimed that several trials were selectively reported, unreported, or the results were improperly calculated by Merck scientists. Lawyers also used evidence of misleading advertising strategies by Merck to discredit the drug company in court.

VIGOR Study (Vioxx GI Outcomes Research)

The VIGOR study began in January 1999 and concluded in March 2000. The study was designed to measure the safety of Vioxx vs. naproxen in arthritis patients, but it did not test the efficacy of Vioxx in treating arthritis.

As early as 4-6 weeks into the trial, the Vioxx group began showing a higher rate of fatal cardiovascular events including heart attack and stroke. By December of 1999, there were twice as many deaths in the Vioxx group. The Data Safety Monitoring Board (DSMB) overseeing the VIGOR study was notified about the deaths, but the data was not immediately interpreted as evidence of the risks of Vioxx. It was hypothesized that the difference in fatalities might have been attributable to the protective anti-platelet effects of naproxen, instead of indicating a complication of Vioxx.

However, in 2006, an investigative report by National Public Radio revealed that the percentage of deaths was steadily rising in the Vioxx group, while remaining almost constant in the naproxen group, which indicated an increased danger with extended use of Vioxx. Additionally, there is no evidence that naproxen’s anti-platelet effects were significant enough to have caused the discrepancy in the VIGOR study.

Perhaps the most ironic finding of the VIGOR study was that the incidence of peptic ulcers was the same in both the Vioxx and naproxen groups. Thus, VIGOR proved that Vioxx had no benefit over traditional pain killers in preventing ulcers.

The results of VIGOR were submitted to the FDA in 2001, and consequently in 2002 a black box warning was ordered for Vioxx packaging, informing consumers of the possible risk of heart attack, stroke, and fatal blood clots.

APPROVe Study (Adenomatous Polyp PRevention On Vioxx)

The APPROVe trial began in 2001 as a 3-year study of the effectiveness of Vioxx vs. placebo in preventing the regrowth of colon and rectal polyps in 2,586 patients with a history of colorectal adenomatous polyps (pre-cancerous masses in the colon and rectum). The study was terminated in September 2004 because of a drastic increase in heart attacks in the Vioxx group. It was due to this study that Merck made the long-standing claim that the risks of Vioxx did not appear until after 18 months of treatment. Merck maintained the 18-month theory until 2008, when the results of several Vioxx studies were scrutinized and it was discovered that as many as 25% of the heart attacks in Vioxx trials occurred within the first month of treatment.

As a direct result of the APPROVe study, Vioxx was withdrawn from the market. The APPROVe study publicized the cardiovascular risks of Vioxx, and it marked the beginning of a string of lawsuits that have cost Merck billions. Because of the drug’s withdrawal, Merck’s handling of Vioxx was placed under intense scrutiny. It was later discovered that the APPROVe study results revealed that taking Vioxx carried even more risk than previously thought.

In 2006, a New England Journal of Medicine article revealed that the APPROVe study results, specifically the number of Vioxx-related deaths during and after the study period, were improperly calculated in order to make the drug appear safer. Some numbers appeared not to have been reported at all.

Merck claimed that numbers were “transposed.” However, later inquiries showed that the numerical discrepancy in the results of the APPROVe trial resulted from the fact that, in the “off-drug period” immediately following the discontinuation of the study, deaths in the placebo group were counted for a year, but deaths in the Vioxx group were counted for only 14 days. During that unreported period, the proportion of deaths in the Vioxx group was much greater than in the control group.

In addition to several audits of Merck’s study results for APPROVe, Merck was investigated for their handling of Vioxx from a “sponsorship” perspective. Merck was accused of utilizing professional ghostwriters to author clinical studies, with the idea of making the results sound more favorable.

At every turn, Merck made decisions and oversights that came back to haunt the drug giant later. The entire medical community scrambled to analyze the Vioxx disaster and answer the big questions: how did it happen, and could it happen again?

Several editorials were published accusing Merck, and the drug industry at large, of corruption and profiteering at the cost of public health. In 2006, Merck made an official statement admitting that it had misinterpreted the results of the APPROVe study, and admitting that the cardiovascular risks of Vioxx appeared to be the same before and after the 18-month benchmark. This statement was expected to have disastrous effects for Merck in its legal battles over Vioxx because now patients who had taken Vioxx for a short time were more likely to file successful lawsuits.

ADVANTAGE Trial (Assessment of Differences between Vioxx and Naproxen To Ascertain Gastrointestinal Tolerability and Effectiveness)

ADVANTAGE was a 12-week trial in 2000 designed to measure Vioxx safety. The results of the 6,000-patient ADVANTAGE study were not publicized until 2005, when all Merck-sponsored clinical trials were under intense scrutiny. In its 2001 report to the FDA on the ADVANTAGE trial, Merck reported a total of eight cardiovascular events. But in 2005, it was discovered that one death had been omitted from the results, bringing the total number of events to 9 — 8 of which were in the Vioxx group. It was later speculated that these results were not made public earlier because they did not support Merck’s claim that Vioxx risks only appeared after 18 months of continuous treatment.

VICTOR Trial (Vioxx in Colorectal Cancer Therapy: Definition of Optimal Regime)

Beginning in 2002, the VICTOR trial was designed to measure the effectiveness of Vioxx vs. placebo in preventing the regrowth of colorectal tumors, in patients who had previously undergone surgery to remove colorectal tumors. The study was terminated prematurely in 2004 because of the worldwide Vioxx withdrawal, and the partial results were never published.

In 2007, an article in The Wall Street Journal included a description of the unpublished VICTOR study. According to Merck’s own records, there were 23 cardiovascular events during the study, 16 of which were in the Vioxx group. Of those 16, 12 occurred during the first 12 months of treatment. These findings came as another blow to Merck’s claim that Vioxx posed no cardiovascular risk until after 18 months of treatment. In response to the publication of these results, Merck announced that they would not make an official statement until the data was re-examined more closely.