On appeal from Superior Court of New Jersey, Law Division, Atlantic County, L-1296-05 and L-3553-05.
The opinion of the court was delivered by: Payne, J.A.D.
NOT FOR PUBLICATION WITHOUT THE APPROVAL OF THE APPELLATE DIVISION
Before Judges Axelrad, Payne and Messano.
Defendant, Merck & Co., Inc., appeals from a $15.7 million judgment, awarding compensatory and punitive damages, as well as attorneys' fees and costs, to plaintiffs, John and Irma McDarby, on claims of product liability and consumer fraud arising from Merck's sale of the prescription drug Vioxx, as well as from a $2.27 million judgment awarding damages of $135 and the remainder as attorneys' fees and costs to plaintiffs, Thomas and Joyce Cona, on claims of consumer fraud arising, likewise, from the sale of Vioxx. The claims of the McDarbys and Conas were tried together. We declined to consolidate Merck's appeals, but scheduled them to be heard back-to-back. This opinion addresses both appeals.
We commence this opinion with a statement of facts that could reasonably have been considered by the jury in support of its verdict. Our factual statement is extended, but we regard it as necessary to place in perspective the issues regarding the applicability of the New Jersey Product Liability Act (PLA), N.J.S.A. 2A:58C-1 to -11, and the New Jersey Consumer Fraud Act (CFA), N.J.S.A. 56:8-1 to -156, that underlie this appeal. The record discloses the tension that existed between Merck's scientists and its marketers and, in plaintiffs' view, the pressure on Merck's employees to preserve market share and concomitant profits arising from the sale of Vioxx - a drug envisioned as re-establishing Merck as preeminent in the field of pharmaceutical development and manufacture - regardless of the cardiovascular risks posed by the drug. The record likewise discloses a spirited defense on behalf of Merck. However, as the result of the verdict in plaintiffs' favor, we do not focus on that defense.
Scientists have known for some time that the enzyme cyclooxygenase (COX) catalyzes the synthesis of prostaglandins, which affect pain and inflammation. Non-steroidal anti-inflammatory drugs (NSAIDs) are a class of compounds including ibuprofen (Advil and Motrin), naproxen (Aleve) and aspirin that exert an analgesic and anti-inflammatory effect by decreasing the production of prostaglandins through the inhibition of COX. For that reason, NSAIDs are widely used in the treatment of acute and chronic pain and inflammation, including that caused by rheumatoid arthritis and osteoarthritis. However, NSAIDs have been found to have a deleterious effect on the gastrointestinal (GI) tract, causing perforations, ulcers, and GI bleeding (collectively, PUBs).
In the early 1990s, scientists learned that prostaglandin synthesis in humans is catalyzed by two forms of cyclooxygenase, cyclooxygenase-1 (COX-1) and cyclooxygenase-2 (COX-2). They postulated that COX-1 functions to protect the gastric mucosa and to promote normal platelet function, whereas COX-2 promotes painful inflammation. Development of a drug that could suppress COX-2, while not affecting COX-1, could be beneficial and potentially lucrative.
At the time of these discoveries, Merck was concerned by the forthcoming loss of patent protection for six of its major drugs, and it was actively seeking replacement products. In 1992, Merck synthesized the substance rofecoxib, later trade-named Vioxx(r), a COX-2 inhibitor that the company posited would have potent analgesic and anti-inflammatory properties without associated GI toxicity. At this time, Pfizer was also actively seeking to develop a COX-2 inhibitor, and competition between the two companies for first entry into the market and an accordingly larger market share was intense.
B. Federal Food and Drug Administration Approval of Vioxx
In order to obtain Federal Food and Drug Administration (FDA) approval, Vioxx was required to undergo Phase I, II and III trials*fn1 designed to demonstrate safety and efficacy in humans for the uses proposed by the manufacturer. By 1995, Merck was actively involved in Phase II studies.
At the time, it was known that two hormones, present in the body, affect blood clotting by causing or preventing aggregation of platelets. Thromboxane acts to induce platelet aggregation and to constrict blood vessels, whereas prostacyclin acts in a reverse fashion. The balance between the two hormones is a factor in preventing thromboses or clots. The actions of these substances had been reported by Merck in its Merck's Manual, which described prostacyclin as the "most potent" anti-clotting substance in the body. However, at trial, plaintiffs demonstrated that this entry, potentially relevant to the risks of taking Vioxx, was absent from the 1999 version of the Manual.
In one of the clinical pharmacology studies conducted during the development of Vioxx, researchers noted that the administration of Vioxx to inhibit COX-2 vastly decreased the excretion of the metabolites of prostacyclin, and thus that it likely inhibited the production of prostacyclin itself.*fn2 Levels of the metabolites of thromboxane remained unchanged. In an article by Garret FitzGerald, the study's chief investigator, and others, received for publication on October 19, 1998, FitzGerald hypothesized that if COX-2 inhibitors suppressed prostacyclin generated within blood vessels without suppressing thromboxane, increased clotting, leading to heart attacks and strokes, would result.
As early as April 13, 1998, Vioxx project team meeting minutes noted the unexpected effect of Vioxx on prostacyclin.
Minutes of Merck's May 12, 1998 project team meeting reflect a May 1998 recommendation by Merck's independent board of scientific advisors to "[b]egin from this point onward to systematically collect data on CV [cardiovascular] events in all clinical trials [for Vioxx . . .] utilizing predefined end points for MCI [myocardial infarction], stroke, TIA [transient ischemic attack], unstable angina etc. To accomplish this task, an adjudication committee*fn3 should be established and follow a formal plan." Such adjudication was commenced.
On November 23, 1998, Merck submitted a new drug application (NDA) for Vioxx to the FDA that included FitzGerald's study and a subsequent analysis of cardiovascular events in then-existing Phase II and partially completed Phase III studies. In its discussion of clinical safety, Merck admitted that "theoretically, there might be a risk for thromboembolic cardiovascular adverse experiences with long-term treatment with a COX-2-specific inhibitor compared to long-term NSAID therapy (where COX-1 inhibition inhibits platelet aggregation)." However, Merck stated that statistical analysis had not disclosed statistically significant differences in thromboembolic cardiovascular adverse experiences, regardless of seriousness, between Vioxx-treated patients and those treated with traditional NSAIDs or placebos. None of the trials submitted to the FDA specifically evaluated cardiovascular safety. Additionally, most were short-term in length and did not evaluate patients at high risk for cardiovascular disease.
The application was reviewed by FDA medical officer Dr. Maria Villalba who, in a report dated May 20, 1999, examined, among other things, the thromboembolic and vascular safety of Vioxx, noting that most of the serious adverse events observed in studies submitted in connection with the NDA were cardiovascular in nature, despite the exclusion of patients with a recent history of myocardial infarction or unstable angina, and of patients with a transient ischemic attack or cardiovascular accident within two years of entry into the study. Patients utilizing cardioprotective doses of aspirin were also excluded. Dr. Villalba noted that "[e]valuation of CV thromboembolic events regardless of seriousness shows a numerically higher incidence of ischemic/thromboembolic events (angina, myocardial infarction, CVA [cardiovascular accident], TIA [transient ischemic attack])" in patients taking Vioxx as compared to those taking a placebo, and that there was "a trend toward an increased incidence in longer trials." However, the doctor determined that it was difficult to reach meaningful conclusions regarding this information because of the small number of events, differences in length of exposure and in dose, and lack of large-scale trials using a high (50 mg) dose of Vioxx. She concluded:
In summary: With the available data, it is impossible to answer with complete certainty whether the risk of cardiovascular and thromboembolic events is increased in patients on rofecoxib. A larger database will be needed to answer this and other safety comparison questions.
Patients who need aspirin for cardiovascular reasons*fn4 should not stop aspirin when taking rofecoxib.
Vioxx was approved by the FDA on May 20, 1999, the same day as Villalba's report, as safe and effective for use in the relief of the signs and symptoms of osteoarthritis, for the management of acute pain in adults, and for treatment of primary dysmenorrhea. The labeling required by the FDA did not contain any warnings or precautions regarding cardiovascular risks.
The FDA's letter informing Merck of its approval of Vioxx stated: "If additional information relating to the safety or effectiveness of this drug becomes available, revision of the labeling may be required."
C. Merck's Product Launch
The sale of Vioxx was launched at a million-dollar two-and- one-half-day launch meeting and party in San Francisco in June 1999. There, David Anstice, Merck's President for the Sale and Marketing of its Human Health Products in North America, described Vioxx as a "superstar" that would make Merck "own" the rheumatology market "once again." He announced that Merck would be distributing seventeen million units of Vioxx as samples between then and the end of a year, stating that the short-run cost was worth the opportunity to win market share. Merck employed its largest-ever sales force for the marketing effort, providing sales incentives and targeting over 10,000 physicians; funded a "Health Education Liaison" program at three million dollars per month; and paid doctors to speak about Vioxx. At trial, plaintiffs' counsel highlighted the money spent on advertising and promotion, and compared it with the absence of any funding for a cardiovascular safety study, which Raymond Gilmartin, Merck's Chief Executive Officer, testified that Merck was not required to perform.
During 1999, Merck compiled a substantial list of influential physicians across the country and their views about Vioxx. A chart introduced at trial indicated that some whose views were adverse to Merck were to be visited by upper management and provided with funding for programs or invited to prestigious meetings in "elegant" national or international locations. A number of other doctors were listed as "neutralized" by offers to participate in clinical trials, speaking engagements, or conferences. The legend "discredit" appeared by the name of one doctor, an advocate for Searle.
In addition to its extensive direct marketing of Vioxx to physicians, throughout the period that Vioxx was on the market, Merck engaged in significant direct-to-consumer marketing efforts, including magazine advertisements and television spots featuring ice skater Dorothy Hamill touting the drug and persons able to engage in leisure activities because of their use of Vioxx.
In mid-1998, prior to the approval of Vioxx by the FDA, Merck commenced development of the protocols for a large-scale blinded study of persons with rheumatoid arthritis, given the acronym VIGOR,*fn5 to test whether Vioxx was associated with fewer gastrointestinal adverse events than a comparator NSAID, naproxen (sold in over-the-counter form as Aleve). Dr. Alise Reicin, who at the time of trial was a vice-president of clinical research at Merck, was a primary drafter of the protocols and the study's clinical monitor. The study utilized doses of 500 mg twice a day for naproxen and, at the FDA's request, a 50-mg dose of Vioxx, which was two times the recommended dose. Use of cardioprotective low-dose aspirin was not permitted. Because of the need of the patients for some form of analgesic, the study did not utilize a placebo. The study was monitored by an independent safety monitoring board that analyzed unblinded data at various points during the study's progress. Between January 1999 and July 1999, approximately 8,000 persons were enrolled in the study, and were divided equally into groups taking Vioxx and naproxen. The primary endpoint for the study was a specified number of clinical upper GI events (gastroduodenal perforation or obstruction, upper gastrointestinal bleeding, and symptomatic gastroduodenal ulcers). However, serious cardiovascular events were also noted and adjudicated for use with other studies in a projected pooled or meta-analysis.
In its November 1999 and December 1999 meetings, the data safety monitoring board discussed the increase in deaths and adverse cardiovascular events that was appearing in patients taking Vioxx over those taking naproxen. Although the board did not recommend discontinuance of the trial as a result, it did recommend development of an analysis plan for adjudicated serious cardiovascular events in the VIGOR trial separate from any other planned analysis of that data. Dr. Reicin responded by providing such a plan and stating that the cutoff date for reporting serious vascular events to Merck would be February 10, 2000 - a date that was maintained in a published report of the cardiovascular evidence obtained in the study, despite later-acquired evidence that suggested a further increase in cardiovascular risk.
At the conclusion of the gastrointestinal portion of the trial on March 9, 2000 (one month after the cardiovascular cutoff), the VIGOR study confirmed that Vioxx and naproxen had similar efficacy against rheumatoid arthritis, and that the use of Vioxx resulted in significantly fewer confirmed adverse gastrointestinal events, as Merck had projected. However, it also demonstrated an alarming four-fold increase in the incidence of non-acute myocardial infarctions. Inclusion of three additional myocardial infarctions, reported shortly after the study's thromboembolic event cut-off date, would have created a five-fold increase.
On March 9, 2000, Dr. Edward Scolnick, the President of Merck's Research Division, wrote an e-mail about the VIGOR data that stated: "The CV events are clearly there." Scolnick continued: "It is a shame but it is a low incidence and it is mechanism based as we worried it was." In a April 12, 2000 e-mail to Dr. Reicin, Dr. Scolnick stated:
I will tell you my worry quotient is high. I actually am in minor agony. What I really want to do is a 10000 vs 10000 patient study in mild-moderate OA [osteoarthritis] Tylenol vs Vioxx with prn [as needed] low dose asa [aspirin] for those judged to need it.
[S]afety first primary endpoint and efficacy secondary or co-primary. WE WILL NOT KNOW FOR SURE WHAT IS GOING ON UNTIL WE DO THIS STUDY. PLEASE THINK HARD ABOUT THE DESIGN BEFORE THE PAC MEETING.
The results of the VIGOR study were reported by Merck to the FDA on March 27, 2000. In a section captioned "Cardiovascular Safety," Merck stated that the VIGOR study "provided an opportunity to determine if the NSAID naproxen, which inhibits platelet aggregation, reduced cardiovascular risk compared with the COX-2 inhibitor rofecoxib, which has no effect on platelet aggregation." Merck then reported: "The overall incidence of serious thromboembolic cardiovascular adverse events was low in both treatment groups. However, the incidence of such events was significantly lower in patients on naproxen compared to rofecoxib." The greatest difference was for non-acute myocardial infarctions, with sixteen for Vioxx and five for naproxen. Additionally, fourteen patients treated with Vioxx sustained strokes, whereas only six of the naproxen-treated patients were thus affected. Merck stated further: "The differences in the incidences of cardiovascular SAEs [significant adverse events] between patients who received rofecoxib and patients who received naproxen was observed consistently between men and women, in patients above and below the age of 65, in patients with and without a history of atherosclerotic cardiovascular disease,*fn6 and in patients with or without classic risk factors for cardiovascular disease."
Merck continued its analysis by stating that "VIGOR is the only study to demonstrate a difference in the incidence of serious cardiovascular adverse events in patients treated with rofecoxib compared with another treatment (placebo or NSAID comparator)" and was inconsistent with previous results of Phase II osteoarthritis studies, which showed identical rates of these events in patients on Vioxx and on NSAID comparitors. Merck explained the VIGOR results to the FDA by stating that: "Non-specific COX-1/COX-2 inhibitors such as naproxen may have cardioprotective effects through COX-1 mediated inhibition of platelet aggregation. The longer duration of therapy with naproxen in VIGOR and the size of the trial may have provided a sufficient sample size and period of observation to demonstrate the cardiovascular protective effects of naproxen." Alternatively, Merck noted that therapy with COX-2 selective inhibitors had been shown to cause "moderate" reductions in the synthesis of prostacyclin, a platelet aggregation inhibitor, without COX-1 mediated inhibition of platelet aggregation. "The resulting imbalance could theoretically have mildly proaggregatory platelet effects" that were noticeable in rheumatoid arthritis patients at higher risk for thrombotic events. As a consequence of the VIGOR findings, Merck signaled to the FDA its intent to amend its ongoing trials so as to allow low dose aspirin treatment for patients who may be at risk for cardiovascular events.
On the same day as its submission to the FDA, March 27, 2000, Merck also issued a news release on VIGOR, stressing the gastrointestinal safety of Vioxx by proclaiming that: "Among patients treated with Vioxx, there was a significantly reduced incidence of serious gastrointestinal events compared to patients treated with naproxen." Merck also reported the cardiovascular results of the study, but stated that there were "significantly fewer thromboembolic events" in patients taking naproxen, which it stated was "consistent with naproxen's ability to block platelet aggregation." As a consequence, the news release stated, investigators were being notified to permit the use of low-dose aspirin when appropriate.
Evidence at trial demonstrated that before the announcement was made, Merck's scientists had been unable to locate appropriately focused studies that supported its theory that naproxen had such a pronounced cardioprotective effect, despite the fact that the drug had been on the market for twenty years. Merck did not admit to the possibility that Vioxx was increasing cardiovascular risks.
E. Merck's Supplemental New Drug Application Based on VIGOR
On June 29, 2000, Merck submitted a supplemental new drug application to add the VIGOR results to the Vioxx label. Primarily, Merck sought to disclose that the results of the VIGOR study had provided "conclusive evidence of the improved GI safety of rofecoxib compared with a nonselective NSAID, naproxen." However, in its transmittal letter, Merck also stated: "Other findings in the VIGOR study reinforce the need for aspirin therapy in patients where cardio-protective use of low dose aspirin is indicated."
On August 7, 2000, Merck requested expedited review of its supplemental new drug application, stating:
We believe that the results of the VIGOR trial establish a significant GI safety advantage for rofecoxib over non-selective NSAIDs which is not conveyed in the currently approved product labeling for this drug. Therefore, MRL [Merck Research Laboratories] is concerned that a standard review classification will delay the availability of this important safety information to prescribers and delay our dissemination of this information in a form consistent with the Agency's appraisal of the data.
However, the FDA denied the request because the maker of Celebrex had submitted a similar supplemental new drug application for label changes, which caused the FDA to decide to review both with a public advisory committee that included outside experts.
On October 13, 2000, Merck submitted a safety update report to the FDA, which included data on the eleven additional patients in the VIGOR study who experienced cardiovascular serious adverse experiences eligible for adjudication that were reported after the pre-specified cut-off date. The adjudicated data disclosed three confirmed myocardial infarctions and one confirmed peripheral venous thrombosis on rofecoxib and one confirmed ischemic cerebrovascular accident on naproxen. However, Merck stated: "Inclusion of these patients in the analysis did not meaningfully alter the findings or conclusions of the study." An accompanying table indicated a substantial difference in the relative risk for all thrombotic events among users of Vioxx and naproxen. Divergence between the two groups commenced at one month. The incidence of confirmed acute myocardial infarctions rose from 0.4% to 0.5% in patients treated with Vioxx.
Merck proposed a label change to reflect the VIGOR findings that included the following:
In this study, in order not to confound the analysis of PUBs [perforation, ulcers, bleeding], patients were not permitted to use concomitant aspirin or other anti-platelet drugs. . . . The incidence of confirmed acute myocardial infarction was 0.4% 0.5% in patients treated with VIOXX 50 mg daily and 0.1% in patients treated with naproxen 500 mg twice daily . . . . This is consistent with the known anti-platelet effects of naproxen.
Merck also sought to include language stating that when the four percent of patients for whom aspirin therapy was indicated were removed from the study, the incidence of confirmed acute myocardial infarction was 0.2% 0.3% in patients treated with VIOXX 50 mg daily and 0.1% in patients treated with naproxen. . . . In other controlled clinical trials, spontaneous reports of these cardiovascular events were similar between VIOXX and nonselective NSAID comparators (ibuprofen, diclofenac and nabumetone). VIOXX is not a substitute for aspirin for cardiovascular prophylaxis.
Merck proposed adding a precaution that Vioxx lacked an anti-platelet effect that could substitute for aspirin, and thus that "[p]atients who require low dose aspirin therapy for cardiovascular prophylaxis should continue on aspirin during therapy with VIOXX."
The results of Merck's VIGOR trial were published in the New England Journal of Medicine on November 23, 2000.*fn7 However, the article omitted the adjudicated myocardial infarctions reported after the study's end date, and thus reported that the incidence of myocardial infarction was 0.1% in the naproxen group and 0.4% in the Vioxx group. Again, the difference was cast as a decrease among the naproxen-treated group, rather than an increase among the Vioxx-treated group. As in the labeling proposed by Merck, the authors of the article attributed the difference in rates of myocardial infarction primarily to the incidence of heart attacks among the four percent of patients who should have been taking cardioprotective doses of aspirin, but were not. After citing a meta-analysis of 7,535 patients comparing Vioxx with placebo and other NSAIDs (diclofenac, ibuprofen, and nabumetone) that revealed similar rates of myocardial infarctions in all groups, the article stated that "our results are consistent with the theory that naproxen has a coronary protective effect and highlight the fact that rofecoxib does not provide this type of protection owing to its selective inhibition of cyclooxygenase-2 at its therapeutic doses and at higher doses. The finding that naproxen therapy was associated with a lower rate of myocardial infarction needs further confirmation in larger studies."
In an editorial dated December 29, 2005,*fn8 published in the New England Journal of Medicine, the authors noted that "Three myocardial infarctions, all in the rofecoxib group, were not included in the data submitted to the Journal." Although initially thought to have been unknown to the study's authors at the time of publication, [i]t now appears . . . that at least two of the authors knew about the three additional myocardial infarctions at least two weeks before the authors submitted the first of two revisions and 4 1/2 months before publication of the article.
Lack of inclusion of the three events resulted in an understatement of the difference in risk of myocardial infarction between the rofecoxib and naproxen groups (presented in the article as a reduction in the risk with naproxen but shown here as an increase in the risk with rofecoxib). It also resulted in the misleading conclusion that there was a difference in the risk of myocardial infarction between the aspirin indicated and aspirin not indicated groups. Plaintiff's expert, Dr. Topol,*fn9 criticized the VIGOR article for advocating the "naproxen hypothesis" in the absence of data from any other study to support it, let alone to show that its effect was strong enough to account for the entire differential in cardiovascular event rates; particularly, since aspirin was known to cause only a twenty-five percent reduction in heart attacks. He testified that naproxen had been available for twenty years, and if a cardioprotective effect existed, it should have been noted. Additionally, Dr. Topol observed that the VIGOR study had been conducted on patients with rheumatoid arthritis, but without heart disease. He speculated that the thrombotic effects could be worse in an undifferentiated population suffering from osteoarthritis. Dr. Topol further criticized the authors of the published VIGOR study for their failure to include a chart, provided to the FDA, that showed the divergence between Vioxx and naproxen in terms of heart attacks and serious thrombolic events commencing at four to six weeks.
Additionally, Dr. Topol noted that an excess of serious cardiovascular events was found in two other studies comparing Vioxx with NSAIDs: Protocol 090, which used a low dose (12.5 mg) of Vioxx for only six weeks on osteoarthritis patients and ADVANTAGE,*fn10 which was a twelve-week trial using a 25-mg dose of Vioxx in comparison to a 1,000 mg naproxen dose in osteoarthritis patients. Dr. Topol also cited a 2004 study by Peter Juni that demonstrated that a progressive meta-analysis of the combined patient populations of the VIGOR and the 090 study would have disclosed a statistically significant cardiovascular risk for Vioxx earlier than Merck recognized.
Plaintiff's expert, Dr. Krumholz, a cardiologist, epidemiologist, and a member of the faculty of the Yale Medical School, provided similar testimony, concluding that Merck should have responded to the VIGOR findings by adding a cardiovascular safety warning to the label, even though the VIGOR results were not definitive, because the five-fold difference between Vioxx and naproxen in the rate of heart attacks was "important and consequential" to the decision whether or not to take Vioxx. Additionally, Dr. Krumholz noted that in the ADVANTAGE study, where low dose aspirin was permitted, an excess of myocardial infarctions and sudden deaths, likely from myocardial infarctions, still appeared among patients taking Vioxx. Because aspirin was at least as cardioprotective as naproxen, these results undermined the theory that VIGOR's results were attributable to naproxen's effects.
On November 23, 2000, Merck issued a news release regarding the New England Journal of Medicine article on the VIGOR study, which led with the statement: "In a study of Vioxx published in The New England Journal of Medicine, Vioxx significantly reduced the risk of serious gastrointestinal side effects by half compared to naproxen." The cardiovascular results of the VIGOR study were reported on page three of the news release, which again noted that "significantly fewer heart attacks were seen in patients taking naproxen" and attributed the lowered incidence to naproxen's cardioprotective effect, which was claimed to be "similar to low-dose aspirin." It was noted that "[p]atients taking low-dose aspirin did not participate in VIGOR."
On December 8, 2000, FDA medical officer, Dr. Shari Targum, issued her medical review of the cardiovascular safety of Vioxx based upon VIGOR, two other protocols,*fn11 clinical trial data and prior FDA reviews. She concluded with respect to VIGOR that "there is an increased risk of cardiovascular thrombotic events, particularly myocardial infarction, in the rofecoxib group compared with the naproxen group." However, she observed that "[m]ore difficult is the question of a safety signal for rofecoxib" because of the absence of a placebo group. Although she noted that Merck had claimed that the difference in myocardial infarctions between the Vioxx and naproxen groups was primarily the result of the antiplatelet effects of naproxen, "[t]his hypothesis is not supported by any prospective placebo-controlled trials with naproxen. One can further argue that, no matter what the attribution, the results (from a cardiovascular standpoint) are favorable for naproxen." Dr. Targum rejected Merck's claims that the majority of the cardiovascular events in the VIGOR study occurred in those patients who should have been on aspirin, finding that "[t]he VIGOR data are consistent (i.e., increased events in the rofecoxib group) even in patients who did not fall into the 'aspirin-indicated' subgroup." She dismissed the theory that the results occurred because patients with rheumatoid arthritis were at an increased risk for cardiovascular events stating that, "one is still faced with the difference in cardiovascular events between rofecoxib and naproxen." She observed that, given that premise, "could one not extend this argument to any patient at increased risk of cardiovascular events?" Finally, Dr. Targum rejected results of other studies involving osteoarthritis and Alzheimers disease patients because the dose of Vioxx and length of exposure had not been stated, and the cardiovascular events were not adjudicated.
On December 28, 2000, the FDA asked Merck for a cardiovascular meta-analysis of studies lasting six months or longer that compared Vioxx to placebo, naproxen and other NSAIDs and with separate treatment of Vioxx at 12.5 mg, 25 mg, and 50 mg. It also sought a meta-analysis of the "composite of all active NJSAID comparators" for the most serious cardiothrombic events. A response was provided on January 8, 2001, which concluded that the risk of sustaining a thrombotic cardiovascular event was similar in patients treated with rofecoxib, placebo and non-naproxen non-selective NSAIDs that lack potent inhibition of platelet function. The risk of sustaining a thrombotic cardiovascular event was reduced in patients treated with naproxen as compared to Vioxx. However, Merck again attributed the reduction with naproxen as "likely due to its ability to maintain near maximal inhibition of platelet function throughout its dosing interval." Significantly, the incidence of heart attacks was elevated both when compared to naproxen and to other non-selective NSAIDs, but that fact was not discussed by Merck.
Over the next sixteen months, various events occurred of relevance to the labeling that Merck had proposed. In early February 2001, the Arthritis Advisory Committee convened by the FDA met to discuss, on succeeding days, the VIGOR trials with Vioxx and the CLASS trials with Celebrex. On January 31, 2001, just prior to the Committee's meeting with Merck's representatives on February 8, Dr. Scolnick e-mailed Gilmartin and Anstice, stating in part:
On Monday, I will show you the essence, an update, of the data that supports Vioxx is safe in the CV sense. But I want to point out to all of you at one time that 1. there is no way to prove that in patients with rheumatoid arthritis that ALL of the difference between Vioxx and naproxen is due to the benefit of naproxen. IT IS IMPOSSIBLE TO PROVE THIS; IT IS IMPOSSIBLE TO KNOW THIS WITH CERTAINTY. It is likely if not certain that our label will state the data from VIGOR. It is even likely that words will be used to say that it is not clear if the effect is purely d[ue] to a protective effect of naproxen in this RA [rheumatoid arthritis] patient population.
When the study results came out last year, this fact was patently clear. Since then we have reduced the uncertainty to this very salient point. But it is impossible to dismiss the point. The FDA will NEVER allow it to be fully dismissed. Ther[e] will be great adverse publicity at the meeting. . . . In any case, we need to face the reality of the situation and manage it. Knowing what is about to happen, managing the short term fall out, and facing and managing any longer term consequences.
However, in a February 5, 2001 e-mail to Dr. Reicin and others who would be presenting Merck's position to the FDA, Dr. Scolnick expressed relief about the results, stating: "We all worried to death about the CV events last Spring. Merck is of course always an issue. But I was sick at the thought we might be doing harm to patients. . . . With all the data now available I am no longer worried." And after the meeting, Dr. Scolnick sent his congratulations, stating: "I bit my nails all day. You were FANTASTIC. You made them look like grade d high school students and you won big huge and completely."
After hearing presentations from Merck representatives, including Dr. Reicin, the advisory committee recommended the inclusion of the VIGOR data, including its cardiovascular component, on the label. It thought the fact that Vioxx was not an inhibitor of platelet clumping should be highlighted, and it suggested further research on the cause of the adverse thromboembolic findings, the significance of which was unclear.
On February 8, Merck issued a press release about the advisory committee meeting declaring its belief that the data presented at the meeting "support the excellent safety profile of Vioxx." It then collectively referred to the pre-approval clinical trials and the ongoing Alzheimer's and ADVANTAGE trials as "clinical trials with Vioxx [at] 12.5 mg, 25 mg and 50 mg in 30,000 patients," and declared that they exhibited "no difference in the incidence of CV events, such as heart attacks, among patients taking Vioxx, other NSAIDs and placebo."
On the following day, Merck sent a bulletin for Vioxx regarding the advisory committee meeting to all sales personnel with responsibility for Vioxx that commenced:
DO NOT INTITIATE DISCUSSIONS ON THE FDA ARTHRITIS ADVISORY COMMITTEE (ADVISORY COMMITTEE) REVIEW OR THE RESULTS OF THE Vioxx(r) GI OUTCOMES RESEARCH (VIGOR) STUDY. YOU MAY RESPOND TO CUSTOMER INQUIRIES ONLY AS OUTLINED BELOW.
The bulletin then instructed sales personnel to "Stay Focused On Efficacy" and, in its summary conclusion stated: "Do not proactively discuss the Advisory Committee Meeting or VIGOR.*fn12
Respond to questions about the study by requesting a PIR [physician information report] and in accordance with the obstacle-handling guide." Physician inquiry twenty-three on the "Obstacle Response Guide" was "I am concerned about the cardiovascular effects of Vioxx." If the inquiry was specific to heart attacks, the sales person was instructed to state:
Doctor, once daily Vioxx has no effect on platelet aggregation, and therefore would not be expected to demonstrate reductions in MI or other CV events. Agents such as low-dose aspirin are routinely prescribed for CV patients for their effect on the inhibition of platelet aggregation. Therefore, once daily Vioxx(r) is not a substitute for aspirin for cardiovascular prophylaxis.
After assuring the physician that Vioxx and aspirin could be taken together, the sales person was instructed to transition back to the positive messages for the drug. If probed further by the physician, the sales person was instructed to offer to submit a physician information request.
An updated proposed label was sent by Merck to the FDA in March 2001 that placed the cardiovascular results of the VIGOR study in a "precautions" section, rather than in a more prominent and more significant "warnings" section. In July 2001, Merck predicted that Vioxx sales for 2002 would be approximately $1.6 to $2.1 billion. Another July long-range planning document projected that Vioxx sales would peak in 2003 before declining, and it stated that if the upcoming cardiovascular labeling were "milder; no prothrombic language," an "upside" estimate of a 25% increase over projected sales for 2006 over baseline projections could be expected; that the "base" earnings assumed that cardiovascular effects would be detailed in the "precautions" section of the labels of all COX-2 inhibitors; and that if cardiovascular effects were placed in the "warnings" section, a 50% decrease in projected sales in 2006 would result.
On March 30, 2001, FDA reviewer Villalba issued a review of Vioxx, reporting that the VIGOR study had revealed a relative risk of developing serious CV/thrombotic events that was more than twice that in the Vioxx group as compared to the naproxen group, mainly because of the difference in the number of myocardial infarctions: 20 with Vioxx and 4 with naproxen. Significantly, she also observed that Merck's proposed theory of the cardioprotective effect of naproxen (58% decrease in the risk of serious CV thrombotic events over a nine-month period) "exceeds that reported in the literature for an anti-platelet agent in a primary or secondary prevention setting." She additionally noted that Merck had completed a twelve-week, 5,500 patient safety study known as ADVANTAGE comparing 25 mg Vioxx with 500 mg naproxen twice daily in a population that did not exclude the use of low-dose aspirin. Although the study had been completed in March 2000, it had not been submitted to the FDA for review, but had been requested. Villalba was interested in determining whether the ADVANTAGE study, which had permitted the use of low-dose aspirin, similarly disclosed a higher cardiovascular risk from use of Vioxx.
An April 6, 2001 letter from the FDA, stating that Merck's supplemental new drug application was "approvable," but requiring the ADVANTAGE data and a safety update report, was met with dismay by Dr. Scolnick, who feared that "our competitor would get a better label now, while Merck was required to provide additional data." In an e-mail dated April 5, 2001, he stated:
I am going in 2 weeks to an FDA Science Board I am on and I have been asked to give a talk on how they can keep their scientists up to date. I have already told them I think their review system is an anachronism because they cannot possibly keep up with science given their hiring constraints. I will be making quite radical suggestions. They have said they will allow me to speak on them.
Dr. Scolnick stated further that, if necessary, he would go to contacts he had made in the Department of Health and Human Services, regarding the labeling situation. In an e-mail of April 9, Dr. Scolnick stated:
I think giving them Advantage was not wise.
The Alzheimer's data vs placebo is helpful. Advantage is not, numbers are too small.
They will data dredge as they did on original submission and we will end up with bad labeling. If they are data dredging Advantage I would argue for giving them the safety data in Alzheimer's since it is much more supportive.
On May 22, 2001, the New York Times published an article on the first page of its business section, captioned "Doubts Are Raised on the Safety of 2 Popular Arthritis Drugs," that noted the higher risk of heart attacks among users of Vioxx revealed by the VIGOR study. A May 2001 "Dear Doctor" letter, sent by Merck in response, stressed the safety of the drug, as did a widely-used "CV Card" utilized to overcome what Merck characterized as the "cardiovascular obstacle," that did not include the VIGOR data. Sales personnel were instructed not to leave the CV Card with physicians, but merely to show it to them. Additionally, a "Dodge Ball Vioxx" documents instructed sales representatives how to "dodge" obstacles that included questions about Vioxx's risk for edema, hypertension, and myocardial infarction by use of Merck's "obstacle handler."
An August 28, 2001 form letter, sent by Merck in response to a physician's request for safety information, stated that the cardiovascular event rate was 0.4 or 0.5% depending on the reporting date. However, at trial, Anstice admitted that this figure related only to heart attacks, and that the cardiovascular event rate, including hypertension and other conditions, was 14.6%, and that the 0.4-0.5% figure was "inaccurate."
On August 22, 2001, the Journal of the American Medical Association published a "special communication" authored by Drs. Debabrata Mukherjee and Steven Nissen, both members of the FDA's Arthritis Steering Committee, and by Dr. Topol,*fn13 that evaluated Merck's VIGOR study and Pfizer's CLASS study involving Celebrex, concluding that "[t]he available data raise a cautionary flag about the risk of cardiovascular events with COX-2 inhibitors." The authors continued: "Given the remarkable exposure and popularity of this new class of medications, we believe that it is mandatory to conduct a trial specifically assessing cardiovascular risk and benefit of these agents. Until then, we urge caution in prescribing these agents to patients at risk for cardiovascular morbidity." Shortly after the publication of this article, Merck responded with an August 2001 "Dear Doctor" letter that criticized the data utilized by the article's authors and stated that Merck stood behind the overall and cardiovascular safety profile of Vioxx.
Also in August 2001, FitzGerald published an study in the New England Journal of Medicine*fn14 in which he speculated on the cause of the elevated incidence of major cardiovascular events with the use of Vioxx in the VIGOR trials and the dissimilar results obtained in the CLASS trial and urged additional research in this area. Significantly, FitzGerald stated that: "There is no convincing evidence from epidemiologic studies that NSAIDs, including naproxen, protect against cardiovascular events." FitzGerald urged additional research on the cardiovascular effects of the COX-2 inhibitors.
The need for further study was echoed by Dr. Scolnick in a September 13, 2001 memo that stated, in relevant part,
[Merck Research Laboratories] has just completed its annual planning meeting. As most of you know we reviewed strategy for each franchise . . . . I want to give you a list of the only studies that I regard as ESSENTIAL. Essential means just that ESSENTIAL. Not preferred, not useful, not helpful; ESSENTIAL. . . .
1. For Vioxx: Only the CV outcome study ONLY ESSENTIAL STUDY! [Spelling and punctuation modified.]
While Dr. Scolnick was urging more research, on September 13, 2001, Anstice sent a voice-mail message to Merck's sales force, in response to complaints, that reminded the sales persons that they had received a "cardiovascular letter" and press release in response to the negative article in the New York Times, published in May, and similar material in response to the negative article in the Journal of the American Medical Association, published in August. The voice mail continued:
I can understand why people are confused about the results of VIGOR that showed differences in heart attacks of .1 versus .5 if they don't understand the data. I can even understand why doctors, why Wall Street, why maybe even lawyers might be confused. To understand VIGOR you must understand Naproxen is cardioprotective.
However, four days later, on September 17, 2001, Merck received a warning letter from the FDA that stated, on the basis of its review of promotional audio conferences given on Merck's behalf by a named physician, a press release, and oral representations made by Merck sales representatives to promote Vioxx, the FDA had concluded that Merck's promotional activities and materials were "false, lacking in fair balance, or otherwise misleading in violation of the Federal Food, Drug, and Cosmetic Act (the Act) and applicable regulations. See 21 U.S.C. §§ 331(a) and (b), 352(a),(f), and (n), and 355(a)." Specifically, the letter stated:
You have engaged in a promotional campaign for Vioxx that minimizes the potentially serious cardiovascular findings that were observed in the Vioxx Gastrointestinal Outcomes Research (VIGOR) study, and thus, misrepresents the safety profile for Vioxx. Specifically, your promotional campaign discounts the fact that in the VIGOR study, patients on Vioxx were observed to have a four to five fold increase in myocardial infarctions (MIs) compared to patients on the comparator non-steroidal anti-inflammatory drug (NSAID), Naprosyn (naproxen).
Although the exact reason for the increased rate of MIs observed in the Vioxx treatment group is unknown, your promotional campaign selectively presents the following hypothetical explanation for the observed increase in MIs. You assert that Vioxx does not increase the risk of MIs and that the VIGOR finding is consistent with naproxen's ability to block platelet aggregation like aspirin. That is a possible explanation, but you fail to disclose that your explanation is hypothetical, has not been ...