Mcdarby.merck.exc.Njad.2008

Page 1401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***

JOHN McDARBY and IRMA McDARBY, husband and wife, Plaintiffs-Respondents, v. MERCK & CO., INC. Defendant-Appellant. THOMAS CONA and

JOYCE CONA, Plaintiffs-Respondents, v. MERCK & CO., INC. Defendant-Appellant.

DOCKET NO. A-0076-07T1, A-0077-07T1

SUPERIOR COURT OF NEW JERSEY, APPELLATE DIVISION

401 N.J. Super. 10; 949 A.2d 223; 2008 N.J. Super. LEXIS 116

January 16, 2008, Argued May 29, 2008, Decided

SUBSEQUENT HISTORY: [***1] Approved for Publication May 29, 2008.Certification granted by, in part McDarby v. Merck &Co., Inc., 196 N.J. 597, 960 A.2d 393, 2008 N.J. LEXIS1540 (2008)Certification denied by McDarby v. Merck & Co., Inc.,196 N.J. 597, 960 A.2d 393, 2008 N.J. LEXIS 1591(2008)Appeal dismissed by, Review improvidently allowed byJohn McDarby v. Merck, 2009 N.J. LEXIS 1003 (N.J.,May 7, 2009)

PRIOR HISTORY: On appeal from Superior Court ofNew Jersey, Law Division, Atlantic County, L-1296-05and L-3553-05.

DISPOSITION: Affirmed in part and reversed in part.

COUNSEL: Douglas S. Eakeley argued the cause forappellant in both cases (Lowenstein Sandler, attorneys;Mr. Eakeley, Michael Dore and Alan S. Modlinger, onthe brief).

Ellen Relkin argued the cause for respondents John andIrma McDarby (Weitz & Luxenberg, attorneys; GeorgeW. Conk, of counsel, Ms. Relkin and Stephen J. Riegel,on the brief).

Evan M. Janush (The Lanier Law Firm) of the New Yorkbar, admitted pro hac vice, argued the cause forrespondents Thomas and Joyce Cona (The Lanier LawFirm, attorneys; W. Mark Lanier, Mr. Janush, andRichard D. Meadow, on the brief).

JUDGES: Before Judges AXELRAD, PAYNE and

MESSANO. The opinion of the Court was delivered byPAYNE, J.A.D.

OPINION BY: PAYNE

OPINION

[*20] [**229] The opinion of the Court wasdelivered by

PAYNE, J.A.D.

Defendant, Merck & Co., Inc., appeals from a $ 15.7million judgment, awarding compensatory and punitivedamages, as well as attorneys' fees and costs, toplaintiffs, John and Irma McDarby, on claims of productliability and consumer fraud arising from Merck's sale ofthe prescription drug Vioxx, as well as from a $ 2.27million judgment [***2] awarding damages of $ 135 andthe remainder as attorneys' fees and costs to plaintiffs,Thomas and Joyce Cona, on claims of consumer fraudarising, likewise, from the sale of Vioxx. The claims ofthe McDarbys and Conas were tried together. Wedeclined to consolidate Merck's appeals, but scheduledthem to be heard back-to-back. This opinion addressesboth appeals.

I.

We commence this opinion with a statement of factsthat could reasonably have been considered by the jury insupport of its verdict. Our factual statement is extended,but we regard it as [*21] necessary to place inperspective the issues regarding the applicability of theNew Jersey Product Liability Act (PLA), N.J.S.A.2A:58C-1 to -11, and the New Jersey Consumer FraudAct (CFA), N.J.S.A. 56:8-1 to -156, that underlie thisappeal. The record discloses the tension that existed

Page 2401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***between Merck's scientists and its marketers and, inplaintiffs' view, the pressure on Merck's employees topreserve market share and concomitant profits arisingfrom the sale of Vioxx -- a drug envisioned as re-establishing Merck as preeminent in the field ofpharmaceutical development and manufacture --regardless of the cardiovascular risks posed by the [***3] drug. The record likewise discloses a spiriteddefense on behalf of Merck. However, as the result of theverdict in plaintiffs' favor, we do not focus on thatdefense.

A. Background

Scientists have known for some time that theenzyme cyclooxygenase (COX) catalyzes [**230] thesynthesis of prostaglandins, which affect pain andinflammation. Non-steroidal anti-inflammatory drugs(NSAIDs) are a class of compounds including ibuprofen(Advil and Motrin), naproxen (Aleve) and aspirin thatexert an analgesic and anti-inflammatory effect bydecreasing the production of prostaglandins through theinhibition of COX. For that reason, NSAIDs are widelyused in the treatment of acute and chronic pain andinflammation, including that caused by rheumatoidarthritis and osteoarthritis. However, NSAIDs have beenfound to have a deleterious effect on the gastrointestinal(GI) tract, causing perforations, ulcers, and GI bleeding(collectively, PUBs).

In the early 1990s, scientists learned thatprostaglandin synthesis in humans is catalyzed by twoforms of cyclooxygenase, cyclooxygenase-1 (COX-1)and cyclooxygenase-2 (COX-2). They postulated thatCOX-1 functions to protect the gastric mucosa and topromote normal [***4] platelet function, whereas COX-2 promotes painful inflammation. Development of a drugthat could suppress [*22] COX-2, while not affectingCOX-1, could be beneficial and potentially lucrative.

At the time of these discoveries, Merck wasconcerned by the forthcoming loss of patent protectionfor six of its major drugs, and it was actively seekingreplacement products. In 1992, Merck synthesized thesubstance rofecoxib, later trade-named Vioxx®, a COX-2 inhibitor that the company posited would have potentanalgesic and anti-inflammatory properties withoutassociated GI toxicity. At this time, Pfizer was alsoactively seeking to develop a COX-2 inhibitor, andcompetition between the two companies for first entryinto the market and an accordingly larger market sharewas intense.

B. Federal Food and Drug Administration Approval ofVioxx

In order to obtain Federal Food and DrugAdministration (FDA) approval, Vioxx was required toundergo Phase I, II and III trials 1 designed to

demonstrate safety and efficacy in humans for the usesproposed by the manufacturer. By 1995, Merck wasactively involved in Phase II studies.

1 Phase I trials normally involve a small group(20 to 80) of healthy volunteers [***5] who areutilized to assess the safety of a investigationalnew drug (IND) over a range of doses. Onceinitial safety has been confirmed in Phase I trials,Phase II trials are performed on larger groups (20to 300) and are utilized to assess how well thedrug functions. Phase II may be divided intoPhase IIA, which studies dosing and Phase IIB,which studies efficacy. Phase III studies usuallyinvolve randomized controlled multicenter trialson large patient groups, which may continuewhile a new drug application is pending beforethe FDA. Such trials may also be used todemonstrate that the drug works for additionalpatients or conditions beyond the original use forwhich the drug was approved for marketing, toobtain additional safety data or to supportmarketing claims. Phase IV trials are conductedto provide safety surveillance and technicalsupport after a drug is approved for sale.

At the time, it was known that two hormones,present in the body, affect blood clotting by causing orpreventing aggregation of platelets. Thromboxane acts toinduce platelet aggregation and to constrict bloodvessels, whereas prostacyclin acts in a reverse fashion.The balance between the two hormones [***6] is afactor in [*23] preventing thromboses or clots. Theactions of these substances had been reported by Merckin its Merck's Manual, which described prostacyclin asthe "most potent" anti-clotting substance in the body.However, at trial, plaintiffs demonstrated that this entry,potentially relevant to the [**231] risks of taking Vioxx,was absent from the 1999 version of the Manual.

In one of the clinical pharmacology studiesconducted during the development of Vioxx, researchersnoted that the administration of Vioxx to inhibit COX-2vastly decreased the excretion of the metabolites ofprostacyclin, and thus that it likely inhibited theproduction of prostacyclin itself. 2 Levels of themetabolites of thromboxane remained unchanged. In anarticle by Garret FitzGerald, the study's chiefinvestigator, and others, received for publication onOctober 19, 1998, FitzGerald hypothesized that if COX-2inhibitors suppressed prostacyclin generated within bloodvessels without suppressing thromboxane, increasedclotting, leading to heart attacks and strokes, wouldresult.

2 Merck's senior scientist, Dr. Nancy Santanello,testified that "Vioxx apparently has the effect oflowering prostacyclin such that there's [***7]

Page 3401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***less prostacyclin. I believe it is about a 50 percentdecrease."

As early as April 13, 1998, Vioxx project teammeeting minutes noted the unexpected effect of Vioxx onprostacyclin. Minutes of Merck's May 12, 1998 projectteam meeting reflect a May 1998 recommendation byMerck's independent board of scientific advisors to"[b]egin from this point onward to systematically collectdata on CV [cardiovascular] events in all clinical trials[for Vioxx . . .] utilizing predefined end points for MCI[myocardial infarction], stroke, TIA [transient ischemicattack], unstable angina etc. To accomplish this task, anadjudication committee 3 should be established andfollow a formal plan." Such adjudication wascommenced.

3 An adjudication committee examines reportedevents to determine accuracy; for example,whether symptoms reported as angina insteadreflect a heart attack.

[*24] On November 23, 1998, Merck submitted anew drug application (NDA) for Vioxx to the FDA thatincluded FitzGerald's study and a subsequent analysis ofcardiovascular events in then-existing Phase II andpartially completed Phase III studies. In its discussion ofclinical safety, Merck admitted that "theoretically, theremight [***8] be a risk for thromboemboliccardiovascular adverse experiences with long-termtreatment with a COX-2-specific inhibitor compared tolong-term NSAID therapy (where COX-1 inhibitioninhibits platelet aggregation)." However, Merck statedthat statistical analysis had not disclosed statisticallysignificant differences in thromboembolic cardiovascularadverse experiences, regardless of seriousness, betweenVioxx-treated patients and those treated with traditionalNSAIDs or placebos. None of the trials submitted to theFDA specifically evaluated cardiovascular safety.Additionally, most were short-term in length and did notevaluate patients at high risk for cardiovascular disease.

The application was reviewed by FDA medicalofficer Dr. Maria Villalba who, in a report dated May 20,1999, examined, among other things, thethromboembolic and vascular safety of Vioxx, notingthat most of the serious adverse events observed instudies submitted in connection with the NDA werecardiovascular in nature, despite the exclusion of patientswith a recent history of myocardial infarction or unstableangina, and of patients with a transient ischemic attack orcardiovascular accident within two years [***9] of entryinto the study. Patients utilizing cardioprotective doses ofaspirin were also excluded. Dr. Villalba noted that"[e]valuation of CV thromboembolic events regardless ofseriousness shows a numerically higher incidence ofischemic/thromboembolic events (angina, myocardialinfarction, CVA [cardiovascular accident], TIA [transient

ischemic [**232] attack])" in patients taking Vioxx ascompared to those taking a placebo, and that there was "atrend toward an increased incidence in longer trials."However, the doctor determined that it was difficult toreach meaningful conclusions regarding this informationbecause of the small number of events, differences inlength of exposure and in dose, and [*25] lack of large-scale trials using a high (50 mg) dose of Vioxx. Sheconcluded:

In summary: With the available data, itis impossible to answer with completecer ta inty whether the r isk ofcardiovascular and thromboembolicevents is increased in patients onrofecoxib. A larger database will beneeded to answer this and other safetycomparison questions.

Patients who need aspirin forcardiovascular reasons 4 should not stopaspirin when taking rofecoxib.

4 Aspirin is recognized to be a cardioprotective [***10] drug because it prevents platelets fromaggregating. Low dose aspirin is oftenrecommended for patients at risk of heart attack.

Vioxx was approved by the FDA on May 20, 1999,the same day as Villalba's report, as safe and effective foruse in the relief of the signs and symptoms ofosteoarthritis, for the management of acute pain in adults,and for treatment of primary dysmenorrhea. The labelingrequired by the FDA did not contain any warnings orprecautions regarding cardiovascular risks.

The FDA's letter informing Merck of its approval ofVioxx stated: "If additional information relating to thesafety 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-dollartwo-and-one-half-day launch meeting and party in SanFrancisco in June 1999. There, David Anstice, Merck'sPresident for the Sale and Marketing of its Human HealthProducts in North America, described Vioxx as a"superstar" that would make Merck "own" therheumatology market "once again." He announced thatMerck would be distributing seventeen million units ofVioxx as samples between then and the end of a year,stating [***11] that the short-run cost was worth theopportunity to win market share. Merck employed itslargest-ever sales force for the marketing effort,providing sales incentives and targeting over 10,000

Page 4401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***physicians; funded a "Health Education Liaison"program at three million dollars per month; [*26] andpaid doctors to speak about Vioxx. At trial, plaintiffs'counsel highlighted the money spent on advertising andpromotion, and compared it with the absence of anyfunding for a cardiovascular safety study, whichRaymond Gilmartin, Merck's Chief Executive Officer,testified that Merck was not required to perform.

During 1999, Merck compiled a substantial list ofinfluential physicians across the country and their viewsabout Vioxx. A chart introduced at trial indicated thatsome whose views were adverse to Merck were to bevisited by upper management and provided with fundingfor programs or invited to prestigious meetings in"elegant" national or international locations. A number ofother doctors were listed as "neutralized" by offers toparticipate in clinical trials, speaking engagements, orconferences. The legend "discredit" appeared by thename of one doctor, an advocate for Searle.

In addition [***12] to its extensive direct marketingof Vioxx to physicians, throughout the period that Vioxxwas on the market, Merck engaged in significant direct-to-consumer marketing efforts, including magazineadvertisements and television spots [**233] featuring iceskater Dorothy Hamill touting the drug and persons ableto engage in leisure activities because of their use ofVioxx.

D. The VIGOR Study

In mid-1998, prior to the approval of Vioxx by theFDA, Merck commenced development of the protocolsfor a large-scale blinded study of persons withrheumatoid arthritis, given the acronym VIGOR, 5 to testwhether Vioxx was associated with fewer gastrointestinaladverse events than a comparator NSAID, naproxen(sold in over-the-counter form as Aleve). Dr. AliseReicin, who at the time of trial was a vice-president ofclinical research at Merck, was a primary drafter of theprotocols and the study's clinical monitor. The studyutilized doses of 500 mg twice a day for naproxen and, atthe FDA's request, a 50-mg dose of Vioxx, which [*27] was two times the recommended dose. Use ofcardioprotective low-dose aspirin was not permitted.Because of the need of the patients for some form ofanalgesic, the study did not utilize [***13] a placebo.The study was monitored by an independent safetymonitoring board that analyzed unblinded data at variouspoints during the study's progress. Between January 1999and July 1999, approximately 8,000 persons wereenrolled in the study, and were divided equally intogroups taking Vioxx and naproxen. The primary endpointfor the study was a specified number of clinical upper GIevents (gastroduodenal perforation or obstruction, uppergastrointestinal bleeding, and symptomaticgastroduodenal ulcers). However, serious cardiovascular

events were also noted and adjudicated for use with otherstudies in a projected pooled or meta-analysis.

5 VIGOR stands for Vioxx GastrointestinalOutcomes Research study.

In its November 1999 and December 1999 meetings,the data safety monitoring board discussed the increasein deaths and adverse cardiovascular events that wasappearing in patients taking Vioxx over those takingnaproxen. Although the board did not recommenddiscontinuance of the trial as a result, it did recommenddevelopment of an analysis plan for adjudicated seriouscardiovascular events in the VIGOR trial separate fromany other planned analysis of that data. Dr. Reicinresponded [***14] by providing such a plan and statingthat the cutoff date for reporting serious vascular eventsto Merck would be February 10, 2000 -- a date that wasmaintained in a published report of the cardiovascularevidence obtained in the study, despite later-acquiredevidence that suggested a further increase incardiovascular risk.

At the conclusion of the gastrointestinal portion ofthe trial on March 9, 2000 (one month after thecardiovascular cut-off), the VIGOR study confirmed thatVioxx and naproxen had similar efficacy againstrheumatoid arthritis, and that the use of Vioxx resulted insignificantly fewer confirmed adverse gastrointestinalevents, as Merck had projected. However, it alsodemonstrated an alarming four-fold increase in theincidence of non-acute myocardial [*28] infarctions.Inclusion of three additional myocardial infarctions,reported shortly after the study's thromboembolic eventcut-off date, would have created a five-fold increase.

On March 9, 2000, Dr. Edward Scolnick, thePresident of Merck's Research Division, wrote an e-mailabout the VIGOR data that stated: "The CV events areclearly there." Scolnick continued: "It is a shame but it isa low incidence and it is mechanism [***15] based aswe worried it was." In a April 12, 2000 e-mail to Dr.Reicin, Dr. Scolnick stated:

[**234] I will tell you my worryquotient is high. I actually am in minoragony. What I really want to do is a 10000vs 10000 patient study in mild-moderateOA [osteoarthritis] Tylenol vs Vioxx withprn [as needed] low dose asa [aspirin] forthose judged to need it. [S]afety firstprimary endpoint and efficacy secondaryor co-primary. WE WILL NOT KNOWFOR SURE WHAT IS GOING ONUNTIL WE DO THIS STUDY. PLEASETHINK HARD ABOUT THE DESIGNBEFORE THE PAC MEETING.

Page 5401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***

The results of the VIGOR study were reported byMerck to the FDA on March 27, 2000. In a sectioncaptioned "Cardiovascular Safety," Merck stated that theVIGOR study "provided an opportunity to determine ifthe NSAID naproxen, which inhibits plateletaggregation, reduced cardiovascular risk compared withthe COX-2 inhibitor rofecoxib, which has no effect onplatelet aggregation." Merck then reported: "The overallincidence of serious thromboembolic cardiovascularadverse events was low in both treatment groups.However, the incidence of such events was significantlylower in patients on naproxen compared to rofecoxib."The greatest difference was [***16] for non-acutemyocardial infarctions, with sixteen for Vioxx and fivefor naproxen. Additionally, fourteen patients treated withVioxx sustained strokes, whereas only six of thenaproxen-treated patients were thus affected. Merckstated further: "The differences in the incidences ofcardiovascular SAEs [significant adverse events]between patients who received rofecoxib and patientswho received naproxen was observed consistentlybetween men and women, in patients above and belowthe age of 65, in patients with and without a history ofatherosclerotic [*29] cardiovascular disease, 6 and inpatients with or without classic risk factors forcardiovascular disease."

6 Nonetheless, the precaution set forth in the2002 FDA-approved label related only to personswith known ischemic heart disease.

Merck continued its analysis by stating that "VIGORis the only study to demonstrate a difference in theincidence of serious cardiovascular adverse events inpatients treated with rofecoxib compared with anothertreatment (placebo or NSAID comparator)" and wasinconsistent with previous results of Phase IIosteoarthritis studies, which showed identical rates ofthese events in patients on Vioxx and on NSAID [***17] comparitors. Merck explained the VIGORresults to the FDA by stating that: "Non-specific COX-1/COX-2 inhibitors such as naproxen may havecardioprotective effects through COX-1 mediatedinhibition of platelet aggregation. The longer duration oftherapy with naproxen in VIGOR and the size of the trialmay have provided a sufficient sample size and period ofobservation to demonstrate the cardiovascular protectiveeffects of naproxen." Alternatively, Merck noted thattherapy with COX-2 selective inhibitors had been shownto cause "moderate" reductions in the synthesis ofprostacyclin, a platelet aggregation inhibitor, withoutCOX-1 mediated inhibition of platelet aggregation. "Theresulting imbalance could theoretically have mildly pro-aggregatory platelet effects" that were noticeable inrheumatoid arthritis patients at higher risk for thromboticevents. As a consequence of the VIGOR findings, Merck

signaled to the FDA its intent to amend its ongoing trialsso as to allow low dose aspirin treatment for patients whomay be at risk for cardiovascular events.

On the same day as its submission to the FDA,March 27, 2000, Merck also issued a [**235] newsrelease on VIGOR, stressing the gastrointestinal [***18]safety of Vioxx by proclaiming that: "Among patientstreated with Vioxx, there was a significantly reducedincidence of serious gastrointestinal events compared topatients treated with naproxen." Merck also reported thecardiovascular results of the [*30] study, but stated thatthere were "significantly fewer thromboembolic events"in patients taking naproxen, which it stated was"consistent with naproxen's ability to block plateletaggregation." As a consequence, the news release stated,investigators were being notified to permit the use oflow-dose aspirin when appropriate.

Evidence at trial demonstrated that before theannouncement was made, Merck's scientists had beenunable to locate appropriately focused studies thatsupported its theory that naproxen had such apronounced cardioprotective effect, despite the fact thatthe drug had been on the market for twenty years. Merckdid not admit to the possibility that Vioxx was increasingcardiovascular risks.

E. Merck's Supplemental New Drug Application Basedon VIGOR

On June 29, 2000, Merck submitted a supplementalnew drug application to add the VIGOR results to theVioxx label. Primarily, Merck sought to disclose that theresults of the [***19] VIGOR study had provided"conclusive evidence of the improved GI safety ofrofecoxib compared with a nonselective NSAID,naproxen." However, in its transmittal letter, Merck alsostated: "Other findings in the VIGOR study reinforce theneed for aspirin therapy in patients where cardio-protective use of low dose aspirin is indicated."

On August 7, 2000, Merck requested expeditedreview of its supplemental new drug application, stating:

We believe that the results of theVIGOR trial establish a significant GIsafety advantage for rofecoxib over non-selective NSAIDs which is not conveyedin the currently approved product labelingfor this drug. Therefore, MRL [MerckResearch Laboratories] is concerned that astandard review classification will delaythe availability of this important safetyinformation to prescribers and delay ourdissemination of this information in aform consistent with the Agency'sappraisal of the data.

Page 6401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, *** However, the FDA denied the request because the makerof Celebrex had submitted a similar supplemental newdrug application [*31] for label changes, which causedthe FDA to decide to review both with a public advisorycommittee that included outside experts.

On October 13, [***20] 2000, Merck submitted asafety update report to the FDA, which included data onthe eleven additional patients in the VIGOR study whoexperienced cardiovascular serious adverse experienceseligible for adjudication that were reported after the pre-specified cut-off date. The adjudicated data disclosedthree confirmed myocardial infarctions and oneconfirmed peripheral venous thrombosis on rofecoxiband one confirmed ischemic cerebrovascular accident onnaproxen. However, Merck stated: "Inclusion of thesepatients in the analysis did not meaningfully alter thefindings or conclusions of the study." An accompanyingtable indicated a substantial difference in the relative riskfor all thrombotic events among users of Vioxx andnaproxen. Divergence between the two groupscommenced at one month. The incidence of confirmedacute myocardial infarctions rose from 0.4% to 0.5% inpatients treated with Vioxx.

[**236] Merck proposed a label change to reflectthe VIGOR findings that included the following:

In this study, in order not to confoundthe analysis of PUBs [perforation, ulcers,bleeding], patients were not permitted touse concomitant aspirin or other anti-platelet drugs. . . . The incidence of [***21] confirmed acute myocardialinfarction was 0.4% 0.5% in patientstreated with VIOXX 50 mg daily and0.1% in patients treated with naproxen500 mg twice daily . . . . This is consistentwith the known anti-platelet effects ofnaproxen.

Merck also sought to include language stating that whenthe four percent of patients for whom aspirin therapy wasindicated were removed from the study,

the incidence of confirmed acutemyocardial infarction was 0.2% 0.3% inpatients treated with VIOXX 50 mg dailyand 0.1% in patients treated withnaproxen. . . . In other controlled clinicaltrials, spontaneous reports of thesecardiovascular events were similarbetween VIOXX and nonselective NSAIDcomparators (ibuprofen, diclofenac andnabumetone). VIOXX is not a substitutefor aspirin for cardiovascular prophylaxis.

Merck proposed adding a precaution that Vioxx lackedan antiplatelet effect that could substitute for aspirin, andthus that "[p]atients who require low dose aspirin therapyfor cardiovascular [*32] prophylaxis should continue onaspirin during therapy with VIOXX."

The results of Merck's VIGOR trial were publishedin the New England Journal of Medicine on November23, 2000. 7 However, the article [***22] omitted theadjudicated myocardial infarctions reported after thestudy's end date, and thus reported that the incidence ofmyocardial infarction was 0.1% in the naproxen groupand 0.4% in the Vioxx group. Again, the difference wascast 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 thearticle attributed the difference in rates of myocardialinfarction primarily to the incidence of heart attacksamong the four percent of patients who should have beentaking cardioprotective doses of aspirin, but were not.After citing a meta-analysis of 7,535 patients comparingVioxx with placebo and other NSAIDs (diclofenac,ibuprofen, and nabumetone) that revealed similar rates ofmyocardial infarctions in all groups, the article stated that"our results are consistent with the theory that naproxenhas a coronary protective effect and highlight the factthat rofecoxib does not provide this type of protectionowing to its selective inhibition of cyclooxygenase-2 atits therapeutic doses and at higher doses. The finding thatnaproxen therapy was associated with a lower rate ofmyocardial infarction [***23] needs furtherconfirmation in larger studies."

7 Claire Bombardier et al., Comparison ofUpper Gastrointenstinal Toxicity of Rofecoxiband Naproxen in Patients with RheumatoidArthritis, 343 New Eng. J. Med. 1520 (2000).

In an editorial dated December 29, 2005, 8 publishedin the New England Journal of Medicine, the authorsnoted that "Three [*33] myocardial infarctions, all in therofecoxib [**237] group, were not included in the datasubmitted to the Journal." Although initially thought tohave been unknown to the study's authors at the time ofpublication,

[i]t now appears . . . that at least two ofthe authors knew about the threeadditional myocardial infarctions at leasttwo weeks before the authors submittedthe first of two revisions and 4 1/2 monthsbefore publication of the article.

* * *

Lack of inclusion of the three eventsresulted in an understatement of thedifference in risk of myocardial infarction

Page 7401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***between the rofecoxib and naproxengroups (presented in the article as areduction in the risk with naproxen butshown here as an increase in the risk withrofecoxib). It also resulted in themisleading conclusion that there was adifference in the risk of myocardialinfarction between the [***24] aspirinindicated and aspirin not indicated groups.

8 Gregory D. Curfman et al., Expression ofConcern: Bombardier et al., "Comparison ofUpper Gastrointestinal Toxicity of Rofecoxib andNaproxen in Patients with Rheumatoid Arthritis,"New Eng. J. Med. 2000; 343:1520-8, 353 NewEng. J. Med. 2813 (2005). A response by the non-Merck authors of the initial article, ClaireBombardier et al., Response to Expression ofConcern Regarding VIGOR Study, 354 New Eng.J. Med. 1196, appeared on March 16, 2006,followed by Curfman et al., Expression ofConcern Reaffirmed, 354 New Eng. J. Med. 1193(2006).

Plaintiff's expert, Dr. Topol, 9 criticized the VIGORarticle for advocating the "naproxen hypothesis" in theabsence of data from any other study to support it, letalone to show that its effect was strong enough toaccount for the entire differential in cardiovascular eventrates; particularly, since aspirin was known to cause onlya twenty-five percent reduction in heart attacks. Hetestified that naproxen had been available for twentyyears, and if a cardioprotective effect existed, it shouldhave been noted. Additionally, Dr. Topol observed thatthe VIGOR study had been conducted on patients with [***25] rheumatoid arthritis, but without heart disease.He speculated that the thrombotic effects could be worsein an undifferentiated population suffering fromosteoarthritis. Dr. Topol further criticized the authors ofthe published VIGOR [*34] study for their failure toinclude a chart, provided to the FDA, that showed thedivergence between Vioxx and naproxen in terms ofheart attacks and serious thrombolic events commencingat four to six weeks.

9 Dr. Topol's deposition was played to the jury.At the time of the deposition, the doctor wasProvost of the Cleveland Clinic Lerner College ofMedicine, Chief Academic Officer of theCleveland Clinic, and Chair of the Department ofCardiovascular Medicine of the Cleveland Clinic.He is an outspoken critic of the conduct of Merckand the FDA in connection with Vioxx whosecriticisms have been published in journals

including the Journal of the American MedicalAssociation.

Additionally, Dr. Topol noted that an excess ofserious cardiovascular events was found in two otherstudies comparing Vioxx with NSAIDs: Protocol 090,which used a low dose (12.5 mg) of Vioxx for only sixweeks on osteoarthritis patients and ADVANTAGE, 10

which was a twelve-week trial [***26] using a 25-mgdose of Vioxx in comparison to a 1,000 mg naproxendose in osteoarthritis patients. Dr. Topol also cited a2004 study by Peter Juni that demonstrated that aprogressive meta-analysis of the combined patientpopulations of the VIGOR and the 090 study would havedisclosed a statistically significant cardiovascular risk forVioxx earlier than Merck recognized.

10 Assessment of Differences Between VioxxAnd Naproxen To Ascertain GastrointestinalTolerability and Effectiveness.

Plaintiff's expert, Dr. Krumholz, a cardiologist,epidemiologist, and a member of the faculty of the YaleMedical School, provided similar testimony, concludingthat Merck should have responded to the VIGORfindings by adding a cardiovascular safety warning to thelabel, even though the VIGOR results were notdefinitive, because the five-fold difference between[**238] Vioxx and naproxen in the rate of heart attackswas "important and consequential" to the decisionwhether or not to take Vioxx. Additionally, Dr.Krumholz noted that in the ADVANTAGE study, wherelow dose aspirin was permitted, an excess of myocardialinfarctions and sudden deaths, likely from myocardialinfarctions, still appeared among patients [***27] takingVioxx. Because aspirin was at least as cardioprotectiveas naproxen, these results undermined the theory thatVIGOR's results were attributable to naproxen's effects.

On November 23, 2000, Merck issued a newsrelease regarding the New England Journal of Medicinearticle on the VIGOR study, which led with thestatement: "In a study of Vioxx published [*35] in TheNew England Journal of Medicine, Vioxx significantlyreduced the risk of serious gastrointestinal side effects byhalf compared to naproxen." The cardiovascular resultsof the VIGOR study were reported on page three of thenews release, which again noted that "significantly fewerheart attacks were seen in patients taking naproxen" andattributed the lowered incidence to naproxen'scardioprotective effect, which was claimed to be "similarto low-dose aspirin." It was noted that "[p]atients takinglow-dose aspirin did not participate in VIGOR."

On December 8, 2000, FDA medical officer, Dr.Shari Targum, issued her medical review of thecardiovascular safety of Vioxx based upon VIGOR, twoother protocols, 11 clinical trial data and prior FDAreviews. She concluded with respect to VIGOR that

Page 8401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***"there is an increased risk of cardiovascular [***28]thrombotic events, particularly myocardial infarction, inthe rofecoxib group compared with the naproxen group."However, she observed that "[m]ore difficult is thequestion of a safety signal for rofecoxib" because of theabsence of a placebo group. Although she noted thatMerck had claimed that the difference in myocardialinfarctions between the Vioxx and naproxen groups wasprimarily the result of the antiplatelet effects ofnaproxen, "[t]his hypothesis is not supported by anyprospective placebo-controlled trials with naproxen. Onecan further argue that, no matter what the attribution, theresults (from a cardiovascular standpoint) are favorablefor naproxen." Dr. Targum rejected Merck's claims thatthe majority of the cardiovascular events in the VIGORstudy occurred in those patients who should have been onaspirin, finding that "[t]he VIGOR data are consistent(i.e., increased events in the rofecoxib group) even inpatients who did not fall into the 'aspirin-indicated' [*36] subgroup." She dismissed the theory that the resultsoccurred because patients with rheumatoid arthritis wereat an increased risk for cardiovascular events stating that,"one is still faced with the difference [***29] incardiovascular events between rofecoxib and naproxen."She observed that, given that premise, "could one notextend this argument to any patient at increased risk ofcardiovascular events?" Finally, Dr. Targum rejectedresults of other studies involving osteoarthritis andAlzheimers disease patients because the dose of Vioxxand length of exposure had not been stated, and thecardiovascular events were not adjudicated.

11 (1) Study 085, a randomized placebo-controlled study to evaluate the efficacy andsafety of low-dose (12.5 mg) Vioxx against theNSAID nabumetone in patients with osteoarthritisof the knee, with use of low-dose aspirinpermitted and (2) Study 090, a similar study thatdisclosed numerically more myocardialinfarctions in the Vioxx group compared withnabumetone and placebo, as well as morecardiovascular adverse experiences anddiscontinuances due to cardiovascular adverseexperiences in the Vioxx group.

[**239] On December 28, 2000, the FDA askedMerck for a cardiovascular meta-analysis of studieslasting six months or longer that compared Vioxx toplacebo, naproxen and other NSAIDs and with separatetreatment of Vioxx at 12.5 mg, 25 mg, and 50 mg. It alsosought a meta-analysis [***30] of the "composite of allactive NJSAID comparators" for the most seriouscardiothrombic events. A response was provided onJanuary 8, 2001, which concluded that the risk ofsustaining a thrombotic cardiovascular event was similarin patients treated with rofecoxib, placebo and non-naproxen non-selective NSAIDs that lack potent

inhibition of platelet function. The risk of sustaining athrombotic cardiovascular event was reduced in patientstreated with naproxen as compared to Vioxx. However,Merck again attributed the reduction with naproxen as"likely due to its ability to maintain near maximalinhibition of platelet function throughout its dosinginterval." Significantly, the incidence of heart attackswas elevated both when compared to naproxen and toother non-selective NSAIDs, but that fact was notdiscussed by Merck.

F. Revised Labeling

Over the next sixteen months, various eventsoccurred of relevance to the labeling that Merck hadproposed. In early February 2001, the Arthritis AdvisoryCommittee convened by the FDA met to discuss, onsucceeding days, the VIGOR trials with Vioxx [*37] andthe CLASS trials with Celebrex. On January 31, 2001,just prior to the Committee's meeting with [***31]Merck's representatives on February 8, Dr. Scolnick e-mailed Gilmartin and Anstice, stating in part:

On Monday, I will show you theessence, an update, of the data thatsupports Vioxx is safe in the CV sense.But I want to point out to all of you at onetime that 1. there is no way to prove thatin patients with rheumatoid arthritis thatALL of the difference between Vioxx andnaproxen is due to the benefit ofnaproxen. IT IS IMPOSSIBLE TOPROVE THIS; IT IS IMPOSSIBLE TOKNOW THIS WITH CERTAINTY. It islikely if not certain that our label will statethe data from VIGOR. It is even likelythat words will be used to say that it is notclear if the effect is purely d[ue] to aprotective 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 wehave reduced the uncertainty to this verysalient point. But it is impossible todismiss the point. The FDA will NEVERallow it to be fully dismissed. Ther[e] willbe great adverse publicity at the meeting. .. . In any case, we need to face the realityof the situation and manage it. Knowingwhat is about to happen, managing theshort term fall [***32] out, and facingand managing any longer termconsequences.

However, in a February 5, 2001 e-mail to Dr. Reicinand others who would be presenting Merck's position to

Page 9401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***the FDA, Dr. Scolnick expressed relief about the results,stating: "We all worried to death about the CV events lastSpring. Merck is of course always an issue. But I wassick at the thought we might be doing harm to patients. . .. With all the data now available I am no longerworried." And after the meeting, Dr. Scolnick sent hiscongratulations, stating: "I bit my nails all day. You wereFANTASTIC. You made them look like grade d highschool students and you won big huge and completely."

After hearing presentations from Merckrepresentatives, including Dr. Reicin, the advisorycommittee recommended the inclusion of the VIGORdata, including its cardiovascular component, on thelabel. It thought the fact that Vioxx was not an [**240] inhibitor of platelet clumping should be highlighted, andit suggested further research on the cause of the adversethromboembolic findings, the significance of which wasunclear.

On February 8, Merck issued a press release aboutthe advisory committee meeting declaring its belief thatthe data presented [***33] at [*38] the meeting"support the excellent safety profile of Vioxx." It thencollectively referred to the pre-approval clinical trials andthe ongoing Alzheimer's and ADVANTAGE trials as"clinical trials with Vioxx [at] 12.5 mg, 25 mg and 50 mgin 30,000 patients," and declared that they exhibited "nodifference in the incidence of CV events, such as heartattacks, among patients taking Vioxx, other NSAIDs andplacebo."

On the following day, Merck sent a bulletin forVioxx regarding the advisory committee meeting to allsales personnel with responsibility for Vioxx thatcommenced:

DO NOT INITIATE DISCUSSIONSON THE FDA ARTHRITIS ADVISORYC O M M I T T E E ( A D V I S O R YCOMMITTEE) REVIEW OR THERESULTS OF THE Vioxx® GIOUTCOMES RESEARCH (VIGOR)STUDY. YOU MAY RESPOND TOCUSTOMER INQUIRIES ONLY ASOUTLINED BELOW.

The bulletin then instructed sales personnel to "StayFocused On Efficacy" and, in its summary conclusionstated: "Do not proactively discuss the AdvisoryCommittee Meeting or VIGOR. 12 Respond to questionsabout the study by requesting a PIR [physicianinformation report] and in accordance with the obstacle-handling guide." Physician inquiry twenty-three on the"Obstacle Response Guide" was [***34] "I amconcerned about the cardiovascular effects of Vioxx." Ifthe inquiry was specific to heart attacks, the sales person

was instructed to state: Doctor, once daily Vioxx has no effecton platelet aggregation, and thereforewould not be expected to demonstratereductions in MI or other CV events.Agents such as low-dose aspirin areroutinely prescribed for CV patients fortheir effect on the inhibition of plateletaggregation. Therefore, once dailyVioxx® is not a substitute for aspirin forcardiovascular prophylaxis.

After assuring the physician that Vioxx and aspirin couldbe taken together, the sales person was instructed totransition back to the positive messages for the drug. Ifprobed further by the physician, the sales person wasinstructed to offer to submit a physician informationrequest.

12 Anstice testified that the sales people werenot permitted to discuss VIGOR because itsresults were not set forth on Vioxx's label.

[*39] An updated proposed label was sent by Merckto the FDA in March 2001 that placed the cardiovascularresults of the VIGOR study in a "precautions" section,rather than in a more prominent and more significant"warnings" section. In July 2001, Merck predicted [***35] that Vioxx sales for 2002 would beapproximately $ 1.6 to $ 2.1 billion. Another July long-range planning document projected that Vioxx saleswould peak in 2003 before declining, and it stated that ifthe upcoming cardiovascular labeling were "milder; noprothrombic language," an "upside" estimate of a 25%increase over projected sales for 2006 over baselineprojections could be expected; that the "base" earningsassumed that cardiovascular effects would be detailed inthe "precautions" section of the labels of all COX-2inhibitors; and that if cardiovascular effects were placedin the "warnings" [**241] section, a 50% decrease inprojected sales in 2006 would result.

On March 30, 2001, FDA reviewer Villalba issued areview of Vioxx, reporting that the VIGOR study hadrevealed a relative risk of developing seriousCV/thrombotic events that was more than twice that inthe Vioxx group as compared to the naproxen group,mainly because of the difference in the number ofmyocardial infarctions: 20 with Vioxx and 4 withnaproxen. Significantly, she also observed that Merck'sproposed theory of the cardioprotective effect ofnaproxen (58% decrease in the risk of serious CVthrombotic events over a [***36] nine-month period)"exceeds that reported in the literature for an antiplateletagent in a primary or secondary prevention setting." Sheadditionally noted that Merck had completed a twelve-week, 5,500 patient safety study known as

Page 10401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***ADVANTAGE comparing 25 mg Vioxx with 500 mgnaproxen twice daily in a population that did not excludethe use of low-dose aspirin. Although the study had beencompleted in March 2000, it had not been submitted tothe FDA for review, but had been requested. Villalba wasinterested in determining whether the ADVANTAGEstudy, which had permitted the use of low-dose aspirin,similarly disclosed a higher cardiovascular risk from useof Vioxx.

[*40] An April 6, 2001 letter from the FDA, statingthat Merck's supplemental new drug application was"approvable," but requiring the ADVANTAGE data anda safety update report, was met with dismay by Dr.Scolnick, who feared that "our competitor would get abetter label now, while Merck was required to provideadditional data." In an e-mail dated April 5, 2001, hestated:

I am going in 2 weeks to an FDAScience Board I am on and I have beenasked to give a talk on how they can keeptheir scientists up to date. I have alreadytold [***37] them I think their reviewsystem is an anachronism because theycannot possibly keep up with sciencegiven their hiring constraints. I will bemaking quite radical suggestions. Theyhave said they will allow me to speak onthem.

Dr. Scolnick stated further that, if necessary, he would goto contacts he had made in the Department of Health andHuman Services, regarding the labeling situation. In ane-mail of April 9, Dr. Scolnick stated:

I think giving them Advantage was notwise. The Alzheimer's data vs placebo ishelpful. Advantage is not, numbers are toosmall. They will data dredge as they didon original submission and we will end upwith bad labeling. If they are datadredging Advantage I would argue forgiving them the safety data in Alzheimer'ssince it is much more supportive.

On May 22, 2001, the New York Times publishedan article on the first page of its business section,captioned "Doubts Are Raised on the Safety of 2 PopularArthritis Drugs," that noted the higher risk of heartattacks among users of Vioxx revealed by the VIGORstudy. A May 2001 "Dear Doctor" letter, sent by Merckin response, stressed the safety of the drug, as did awidely-used "CV Card" utilized to overcome [***38]what Merck characterized as the "cardiovascularobstacle," that did not include the VIGOR data. Sales

personnel were instructed not to leave the CV Card withphysicians, but merely to show it to them. Additionally, a"Dodge Ball Vioxx" documents instructed salesrepresentatives how to "dodge" obstacles that includedquestions about Vioxx's risk for edema, hypertension,and myocardial infarction by use of Merck's "obstaclehandler."

An August 28, 2001 form letter, sent by Merck inresponse to a physician's request [**242] for safetyinformation, stated that the cardiovascular event rate was0.4 or 0.5% depending on the reporting [*41] date.However, at trial, Anstice admitted that this figurerelated only to heart attacks, and that the cardiovascularevent 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 AmericanMedical Association published a "specialcommunication" authored by Drs. Debabrata Mukherjeeand Steven Nissen, both members of the FDA's ArthritisSteering Committee, and by Dr. Topol, 13 that evaluatedMerck's VIGOR study and Pfizer's CLASS studyinvolving Celebrex, concluding that "[t]he available [***39] data raise a cautionary flag about the risk ofcardiovascular events with COX-2 inhibitors." Theauthors continued: "Given the remarkable exposure andpopularity of this new class of medications, we believethat it is mandatory to conduct a trial specificallyassessing cardiovascular risk and benefit of these agents.Until then, we urge caution in prescribing these agents topatients at risk for cardiovascular morbidity." Shortlyafter the publication of this article, Merck responded withan August 2001 "Dear Doctor" letter that criticized thedata utilized by the article's authors and stated that Merckstood behind the overall and cardiovascular safety profileof Vioxx.

13 Debabrata Mukherjee et al., Risk of Cardio-vascular Events Associated with Selective COX-2Inhibitors, 286 JAMA 954 (2001).

Also in August 2001, FitzGerald published an studyin the New England Journal of Medicine 14 in which hespeculated on the cause of the elevated incidence ofmajor cardiovascular events with the use of Vioxx in theVIGOR trials and the dissimilar results obtained in theCLASS trial and urged additional research in this area.Significantly, FitzGerald stated that: "There is noconvincing evidence [***40] from epidemiologic studiesthat NSAIDs, including naproxen, protect againstcardiovascular events." FitzGerald [*42] urgedadditional research on the cardiovascular effects of theCOX-2 inhibitors.

14 Garret A. FitzGerald & Carlo Patrono, TheCoxibs, Selective Inhibitors of Cyclooxygenase-2,

Page 11401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***345 New Eng. J. Med. 433 (2001).

The need for further study was echoed by Dr.Scolnick in a September 13, 2001 memo that stated, inrelevant part,

[Merck Research Laboratories] has justcompleted its annual planning meeting.As most of you know we reviewedstrategy for each franchise . . . . I want togive you a list of the only studies that Iregard as ESSENTIAL. Essential meansjust that ESSENTIAL. Not preferred, notuseful, not helpful; ESSENTIAL. . . .

1. For Vioxx: Only the CV outcomestudy ONLY ESSENTIAL STUDY!

[Spelling and punctuation modified.]

While Dr. Scolnick was urging more research, onSeptember 13, 2001, Anstice sent a voice-mail messageto Merck's sales force, in response to complaints, thatreminded the sales persons that they had received a"cardiovascular letter" and press release in response tothe negative article in the New York Times, published inMay, and similar material in response [***41] to thenegative article in the Journal of the American MedicalAssociation, published in August. The voice mailcontinued:

I can understand why people areconfused about the results of VIGOR thatshowed differences in heart attacks of .1versus .5 if they don't understand the data.I can even understand why doctors, whyWall Street, why maybe even lawyersmight be confused. To understand [**243] VIGOR you must understand Naproxen iscardioprotective.

However, four days later, on September 17, 2001,Merck received a warning letter from the FDA thatstated, on the basis of its review of promotional audioconferences given on Merck's behalf by a namedphysician, a press release, and oral representations madeby Merck sales representatives to promote Vioxx, theFDA had concluded that Merck's promotional activitiesand materials were "false, lacking in fair balance, orotherwise misleading in violation of the Federal Food,Drug, and Cosmetic Act (the Act) and applicableregulations. 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 thepotentially serious cardiovascular findingsthat [***42] were observed in the VioxxGastrointestinal Outcomes Research(VIGOR) study, and thus, misrepresentsthe safety profile for Vioxx. Specifically,your promotional campaign discounts thefact that in the VIGOR study, patients onVioxx were observed to have a four tofive [*43] fold increase in myocardialinfarctions (MIs) compared to patients onthe comparator non-steroidal anti-inflammatory drug (NSAID), Naprosyn(naproxen).

Although the exact reason for theincreased rate of MIs observed in theVioxx treatment group is unknown, yourpromotional campaign selectively presentsthe following hypothetical explanation forthe observed increase in MIs. You assertthat Vioxx does not increase the risk ofMIs and that the VIGOR finding isconsistent with naproxen's ability to blockplatelet aggregation like aspirin. That is apossible explanation, but you fail todisclose that your explanation ishypothetical, has not been demonstratedby substantial evidence, and that there isanother reasonable explanation, thatVioxx may have pro-thromboticproperties.

* * *

Your minimizing these potential risksand misrepresenting the safety profile forVioxx raise significant public health andsafety concerns. [***43] Yourmisrepresentation of the safety profile forVioxx is particularly troublesome becausewe have previously, in an untitled letter,objected to promotional materials forVioxx that also misrepresented Vioxx'ssafety profile.

The FDA required Merck to cease all violativepromotional activity and to provide a detailed responseby October 1, 2001, including a "Dear HealthcareProvider" letter to correct false or misleading impressionsand information. 15

15 Anstice responded by stating that the FDAhad mistakenly focused on VIGOR, not on areview of all available data which disclosed nosignificant risks for Vioxx when compared to aplacebo. Anstice further stated that Merck's

Page 12401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***agreement with the offending speaker had beenterminated, and he explained Merck's pressreleases as an appropriate response to "media andanalyst activity." He sought to defer the "DearHealthcare Provider" letter until labeling wasfinalized.

Merck reacted to the warning letter by providing, onOctober 1, 2001, further, superseding directions to itssales persons with respect to VIGOR, with the instructionthat "[y]ou may not discuss or respond to any questionsabout VIGOR, except as specifically set forth in this [***44] Bulletin." The document then stated that if askedabout Merck's GI safety study, the results of itsrheumatoid arthritis study, or why Vioxx had a higherrate of heart attacks than naproxen, the sales personshould identify the VIGOR [**244] study, refuse todiscuss its details "[b]ecause the study is not in the label"and to offer to refer the question to Merck's MedicalServices department. The instructions continued: "if youare [*44] asked any other questions about VIGOR by ahealth care professional or a customer, you may notanswer the question. You may respond to unsolicitedquestions only by offering to submit a [physicianinformation request]." If asked about the FDA's warningletter, sales persons were instructed to respond, only:"The Warning Letter is from FDA's Advertising Divisionand relates to Vioxx. We are responding to FDA. Merckcontinues to stand behind the overall and cardiovascularsafety of Vioxx."

Minutes of an FDA regulatory briefing meeting heldon September 21, 2001 disclose Villalba's conclusionthat in VIGOR, "there was no overall safety advantagefor rofecoxib when compared to naproxen," and that"[f]indings in ADVANTAGE, RA safety database andAlzheimer's studies were [***45] not inconsistent withfindings in VIGOR." The minutes also reflect therecommendation that "FDA should strengthen theWARNINGS section of Vioxx, and deemphasize thesafety advantage information in the label. Naproxenshould be used as a comparison in the label."

On October 15, 2001, the FDA sent Merck a draftlabel for Vioxx. In the "Warnings" section of the label,the FDA proposed:

Cardiovascular Disease

VIOXX should be used with cautionin patients at risk of developingcardiovascular thrombotic events such asthose with a history of myocardialinfarction and angina and in patients withpre-existent hypertension and congestiveheart failure.

The risk of developing myocardialinfarction in the VIGOR study was five-

fold higher in patients treated withVIOXX 50-mg (0.5%) as compared topatients treated with naproxen (0.1%)(See Special studies, VIGOR). Thefinding was consistent in a smaller andshorter study using VIOXX 25 mg thatallowed the use of low dose ASA [aspirin](See Special Studies, ADVANTAGE).Prospective, well-powered, long termstudies required to compare the incidenceof serious CV events in patients takingVIOXX versus NSAID comparators otherthan naproxen have not been [***46]performed.

Because of its lack of platelet effect,VIOXX is not a substitute for aspirin forcardiovascular prophylaxis. The impact ofVIOXX on the card iovascularprophylactic benefit of ASA is unknown.(See special studies, Platelets:PRECAUTIONS, Drug Interactions,Aspirin).

In an October 15, 2001 e-mail sent upon receipt ofthe FDA's proposed label, Merck's Dr. Scolnick stated toAnstice:

[*45] David. Be assured we will notaccept this label. If we need to we will askto go to an advisory committee meeting.

Anstice replied:

. . . We knew it would be UGLY and itis. We'll fight back and see where we get.I agree that we should ask for an advisorycommittee if necessary.

To which Scolnick responded:

It is ugly cubed. thye [sic] are bastards.

Merck proposed relocation of the FDA's text to thePrecautions section of the label, and to modify the text tode-emphasize the risk of Vioxx by stating as follows:

Cardiovascular Effects

The following data should be takeninto consideration when prescribingVIOXX in patients at risk of developingcardiovascular thrombotic events.

[**245] The risk of developing a

Page 13401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***serious cardiovascular thrombotic event inthe VIOXX study was significantlydifferent in patients treated [***47] withVIOXX 50 mg once daily as compared topatients treated with naproxen 500 mgtwice daily. This was largely due to thesignificant difference in the incidence ofmyocardial infarction between patientstaking VIOXX 50 mg once daily (0.5%)and naproxen 500 mg twice daily (0.1%).(See CLINICAL STUDIES, SpecialStudies, VIGOR). In other controlledclinical trials, the incidence of all seriouscardiovascular thrombotic events,including myocardial infarction, wassimilar between VIOXX, nonselectiveNSAID comparators ( ibuprofen,diclofenac and nabumetone) and placebo.Prospective, well powered, long termstudies specifically designed to comparethe incidence of serious CV events inpatients taking VIOXX versus NSAIDcomparators have not been performed.

Because of its lack of platelet effects,VIOXX is not a substitute for aspirin forcardiovascular prophylaxis [sentence willappear bold?] (See CLINICAL STUDIES,Specia l S tudies , P la te le ts andPRECAUTIONS, Drug Interactions,Aspirin).

On November 28, 2001, FDA reviewer Villalbaprovided an analysis of Merck's response to the FDA'sapprovable letter, issued on April 7, 2001, whichrequired that Merck submit data from the ADVANTAGEstudy -- [***48] a twelve-week comparison of Vioxx,taken at 25 mg per day, with naproxen, taken at 500 mgtwice per day in 5,400 patients with osteoarthritis -- aSafety Update Report on the long-term follow up ofpatients in Merck's original osteoarthritis program, andsafety data from studies not previously submitted to theFDA. The summary of clinical findings relating to safetywas not favorable to Merck. Dr. Villalba divided heranalysis into three categories: (1) findings that applied to[*46] ADVANTAGE and the VIGOR databases; (2)cardiovascular safety of Vioxx compared to NSAIDsother that naproxen; and (3) cardiovascular safety ofVioxx compared to placebo. With respect to the firstcategory, the doctor found that Vioxx at 25 and 50 mgdoses showed no overall safety advantage over naproxenas measured by total deaths, serious adverse events,hospitalizations, or discontinuations due to adverseevents and common adverse events; that Vioxx wasassociated with a nominally higher incidence ofdiscontinuations due to hypertension, edema and

congestive heart failure-related events; and it wasassociated with a nominally higher cardiovascularthrombotic risk, particularly heart attacks. The doctorfound [***49] no adequate long-term data comparingthe cardiovascular risk of Vioxx to traditional NSAIDsother than naproxen. Finally, she found that existingstudies did not provide adequate evidence that Vioxx hasa cardiovascular safety profile similar to placebo. In thatconnection, she reported that the Alzheimer's studiesdisclosed a higher incidence of cardiovascularthrombotic deaths with Vioxx that with placebo (nine vs.four), and also noted that "although this was an elderlypopulation (mean age 75 years), patients at highcardiovascular risk were not enrolled." Additionally, thedoctor found the trend of excess serious cardiacthrombotic events in the ADVANTAGE study anddiscontinuances resulting from such events wasconsistent with VIGOR and "of concern" becauseADVANTAGE was only a twelve-week study, used alower dose of Vioxx, and permitted the use of aspirin forcardiovascular prophylaxis.

On January 12, 2002, Dr. Wayne Ray, a professor ofpreventive medicine, published an article describing hisobservational and retrospective analysis of Tennessee[**246] Medicaid patients for the years 1987 to 1998,before the widespread use of COX-2 inhibitors. 16 In thearticle, he identified the patients [***50] who had beentaking only aspirin and those who had been taking a non-aspirin [*47] NSAID. He found no indication thatnaproxen had a cardioprotective effect.

16 Wayne A. Ray et al., Non-steroidal Anti-inflammatory Drugs and Risk of SeriousCoronary Heart Disease: An ObservationalCohort Study, 359 Lancet 118 (2002).

While approval of a new label remained pending,Merck was negotiating an agreement with Brigham andWomen's Hospital in Boston to perform thecardiovascular risk study urged by Dr. Scolnick, entitlingit "A Randomized, Double Blind, Parallel, Placebo-Controlled Trial to Evaluate the Cardiovascular Safetyand Efficacy of Rofecoxib on Cardiovascular Events inPatients with Recent Acute Coronary Syndromes" --(VALOR)." By February 13, 2002, a letter of intent wassent by Dr. Alan Nies, Senior Vice-President for ClinicalSciences at Merck Research Laboratories to the HarvardMedical School with respect to the study. Despite Dr.Scolnick's urging, the study was never performed.

The revised label for Vioxx was approved on April11, 2002, two years after the results of the VIGOR studywere known, and a "Dear Doctor" letter substantiallyincorporating the information set forth in the label [***51] was circulated by Merck that same month. Areview of the label demonstrates that Merck successfullyobtained the FDA's consent to use of a revised label that

Page 14401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***contained no mention of cardiovascular risks in the"Warnings" section, but instead, contained a "Precaution"that limited use of Vioxx only among patients "with amedical history of ischemic heart disease" -- patientswhose already-diagnosed coronary artery disease wassymptomatic. However, the label set forth the results ofthe VIGOR trials in detail, and it stated that "the risk ofdeveloping a serious cardiovascular thrombolic eventwas significantly higher in patients treated with Vioxx 50mg once daily (n=45) as compared to patients treatedwith naproxen 500 mg twice daily (n=19)." It did notexpress the results in terms of a lower incidence amongthose taking naproxen, and it did not contain Merck'sthesis that naproxen was cardioprotective. Instead, the"Precautions" section stated:

The significance of the cardiovascularfindings from these 3 studies (VIGOR andtwo placebo-controlled studies) isunknown. Prospective studies specificallydesigned to compare the incidence ofserious CV events in patients takingVioxx versus [***52] NSAIDcomparators or placebo have not beenperformed.

[*48] Because of the lack of plateleteffects, Vioxx is not a substitute foraspirin for cardiovascular prophylaxis.

G. Product Withdrawal

Following approval of the revised label, Vioxxcontinued to be marketed until September 30, 2004,when evidence of adverse cardiovascular events resultingfrom Merck's APPROVe study 17 led to the voluntary[**247] withdrawal of the drug from the market. Duringthe period between FDA approval of a revised label forVioxx in April 2002 and Merck's withdrawal of theproduct, scientists, including Merck's Dr. Reicin,published, in October 2003, a meta-analysis of theclinical trials, VIGOR, and the Alzheimer's trials, 18

concluding that "rofecoxib was not associated withexcess CV thrombotic events compared with eitherplacebo or non-naproxen NSAIDs. Again, naproxenappeared to be the outlier, suggesting a cardioprotectivebenefit of naproxen." The authors concluded additionallythat "among the predominantly elderly, male populationparticipating in Alzheimer trials, both [*49] rofecoxib-and placebo-treated patients had similar rates of CVthrombotic events. The totality of data is not consistentwith an increased CV [***53] risk among patientstaking rofecoxib."

17 The APPROVe study (Adenomatous Polyp

Prevention on Vioxx), for which patientenrollment commenced in February 2000, wasproposed as a three-year trial of Vioxx at 25 mgagainst placebo in patients with a history ofcolorectal adenomas or polyps. The primaryendpoint was whether Vioxx could matchaspirin's known effectiveness in reducing therecurrence of polyps while maintaining GI safety.It was also designed to assess CV safetyprospectively. The study excluded patients whowere expected to need long-term NSAID therapy,those who had experienced significantcardiovascular events or conditions during thepreceding year, or a stroke or transient ischemicattack during the preceding two years. The studywas initially reported as Robert S. Bresalier et al.,Cardiovascular Events Associated with Rofecoxibin a Colorectal Adenoma Chemoprevention Trial,352 New Eng. J. Med. 1092 (2005). The articlestated that the relative risk of a confirmedthrombotic event with Vioxx was 1.92, and thedifference between Vioxx and placebo wasprimarily due to an increase in myocardialinfarctions and strokes. In a correction printed onJuly 13, 2006, statements [***54] that theincreased relative risk became apparent aftereighteen months of treatment and that the eventrates were similar between groups in the firsteighteen months were deleted. Correction, 355New Eng. J. Med. 2.18 Matthew R. Weir, Rhoda S. Sperling, AliseReicin, & Barry J. Gertz, Selective COX-2Inhibition and Cardiovascular Effects: A Reviewof the Rofecoxib Development Program, 146 Am.Heart J. 591 (2003).

However, in an editorial published in The Lancet inAugust 2004, 19 Dr. Topol commented on a studydemonstrating the small protective effect of naproxen(less than half that of aspirin) and concluded as a resultthat the continued commercial availability of Vioxxwithout a black-box warning was "indeed troubling."

19 Eric J. Topol and Gary W. Falk, A Coxib aDay Won't Keep the Doctor Away, 364 Lancet639 (2004).

Additionally, in an article published in The Lancet inNovember 2004, 20 Peter Juni and his co-authorsdemonstrated how the cardiovascular risk of Vioxx couldhave been discovered earlier by appropriate cumulativestatistical meta-analysis. The article concluded:

Our cumulative meta-analysis ofrandomised controlled trials indicates thatan increased risk of myocardial infarction [***55] was evident from 2000 onwards.At the end of 2000, the effect was both

Page 15401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***substantial and unlikely to be a chancefinding.

We found an increased risk ofmyocardial infarction in trials of bothshort and long duration, which is incontrast to the unpublished results fromthe APPROVe trial. Our findings thusindicate that patients are at risk even ifrofecoxib is taken for a few months only.Therefore, the reassuring statement byMerck, that there is no excess risk in thefirst 18 months, is not supported by ourdata. Similarly, we recorded no evidenceto support the notion that rofecoxib'scardiovascular toxicity is dose-dependent.

[Footnotes omitted.]

Additionally, the authors challenged the naproxenhypothesis, concluding:

The possible cardioprotective effect ofnaproxen has also been examined inseveral observational, pharmaco-epidemiological studies. Taken together,the data from these studies indicate that ifa protective effect of naproxen exists, it is[**248] probably small, and, as pointedout earlier, not large enough to explain thefindings of VIGOR.

[Footnotes omitted.]

20 Peter Juni et al., Risk of CardiovascularEvents and Rofecoxib: Cumulative Meta-analysis,364 Lancet 2021 (2004).

[*50] Although [***56] the Juni study wasseverely challenged by Merck at trial, plaintiffs' expert,Dr. Krumholz, spoke approvingly of the article andstated that the authors had used proper statisticaltechniques in reaching their conclusions, which wereconsistent with the FitzGerald hypothesis.

II.

Plaintiffs John McDarby and Thomas Cona bothtook Vioxx for osteoarthritic pain commencing prior tothe FDA's approval of Merck's revised label in April2002.

McDarby 21 sustained a heart attack and fractured hipon April 15, 2004 at the age of 75. He was prescribedVioxx by his family physician, Dr. John Braun, onMarch 21, 2000, as treatment for osteoarthritis in thehands and knee, and he took it daily until his heart attack

on April 15, 2004. Prior to 2000, McDarby had not heardof Vioxx; thereafter, he saw a number of Merck'scommercials for the product on television, whichsolidified his thinking that Vioxx was a "goodprescription." McDarby read the drug's package insert atthe time of his first purchase, but could recall none of thecontents, and did not read the insert thereafter, relying onhis physician to determine whether it was safe. McDarbytestified that he would not have taken the drug if he [***57] had been told it could cause heart attacks. At thetime of his treatment by Dr. Braun, McDarby was adiabetic whose condition was controlled by oralmedicine. He had sustained a brief loss of vision thatmight have resulted from a transitory ischemic attack,and therefore took low-dose aspirin. However, theamount of plaque in his carotid arteries was found to benormal. Additionally, McDarby was "slightly"overweight. Dr. Braun found that he did not suffer fromhypertension.

21 We do not discuss Cona's medical history,since the jury did not accept his claim of physicalinjury as the result of taking Vioxx.

Dr. Braun's videotaped deposition testimony wasplayed for the jury. In it, he confirmed that he had treatedMcDarby in the [*51] period from September 9, 1998 toNovember 18, 2003, and that he had prescribed Vioxx ata 25 mg dose as McDarby had stated. The doctor testifiedthat as a matter of practice, he reads the entire packageinsert for a drug before prescribing it for the first time "soI know what to expect from a drug and who I can use itin, who I can't use it in, if it's contraindicated in a certainpatient population, if it's going to cause risk factors inpatients with renal [***58] insufficiency or heart diseaseor whatever, to get a better understanding of the drug andits . . . side effects." He also discussed Vioxx withMerck's sales representatives, who visited his office atleast twice a week. As one sales representativeacknowledged, the doctor was targeted because of thehigh volume of his prescriptions for pain relievers. Dr.Braun was familiar with the VIGOR study CV results,but he testified that he was told that they wereattributable to naproxen's cardioprotective effects, andthat representatives assured him that Vioxx was safe forpatients with CV risks so long as they continued to takeaspirin. On three to four occasions after the VIGORresults became known, Dr. Braun was also shownMerck's "CV Card," entitled Chemical Profile,Osteoarthritis Studies" that indicated no elevated risk ofheart attack and, according to Dr. Braun, showed Vioxxto be safer than a [**249] placebo. Additionally, Dr.Braun testified to having received and relied uponMerck's May 2001 "Dear Doctor" letter that referred tomedia reports regarding the safety profile of Vioxx and"place[d] the information in the news reports in contextby setting forth the results of Merck's osteoarthritis

Page 16401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***[***59] studies as also summarized on the CV Card." Dr.Braun testified that he understood the letter as"reaffirm[ing] that the drug was safe."

Although the doctor testified additionally that he hadread Dr. Topol's article in the Journal of the AmericanMedical Association, which indicated that VIGOR's CVresults theoretically could be attributed to theprothrombic effect of Vioxx, the antithrombic effect ofnaproxen, or both, he understood the article to besuggesting the need for additional studies, not that use ofVioxx be suspended. Additionally, he was reassured by,Merck's statements in an August 2001 "Dear Doctor"letter that was critical of the [*52] Topol data and statedthat Merck stood by the cardiovascular safety profile ofits drug. Although Dr. Braun understood that Vioxx wasnot cardioprotective, he testified that he was never toldby a Merck representative that use of Vioxx increasedclotting risks.

Dr. Braun testified that if he had been informed ofthe cardiovascular risks of Vioxx, he would not haveprescribed it to McDarby. In this connection, thefollowing exchange occurred:

Q. If you had been told by Merck thatVioxx could increase the risk of a heartattack, would you have prescribed [***60] Vioxx to Mr. McDarby?

THE WITNESS: Of course not.

Q. Why not?

A. My . . . job as a doctor is to try toprevent things from happening, try toprevent strokes, try to prevent heartattacks.

He [McDarby] has one risk factor thatwe know of, which is diabetes. His secondrisk factor is being a male. And his thirdrisk factor is being elderly for havingheart disease. So why would I give himanother risk factor? Why would I give hima thromboembolic drug, a drug thatcaused clots?

That's not my job.

Dr. Braun testified further that, after the April 2002revised label was issued, he understood Vioxx to becontraindicated in patients with ischemic heart disease,and he had in fact stopped prescribing the drug to apatient for whom the use was contraindicated. However,McDarby did not have that condition, and thus theprescription was continued.

III.

Merck has challenged the jury's verdict in favor ofplaintiff McDarby on his product liability claim, arguingfirst that the trial judge failed to give proper effect to thePLA's presumption of adequacy for prescription drugwarnings approved by the FDA and, second, that theFederal Food Drug and Cosmetic Act (FDCA), 21U.S.C.A. §§ 301 to 399, preempts [***61] McDarby'sclaims challenging the adequacy of the FDA-approvedVioxx labels. McDarby responds (1) that state lawimposes a duty upon manufacturers of prescription drugsto warn of the drug's dangers as soon as knowledge ofthose dangers exists; (2) that the trial judge [*53] properly applied the rebuttable presumption of warningadequacy contained in the PLA; and (3) that the judgewas correct in her rulings and instructions that, as amatter of law, Merck had a duty to warn of thecardiovascular risks of Vioxx without seeking FDAapproval. McDarby also argues that the PLA, asapplicable to claims of inadequate warnings bypharmaceutical manufacturers, [**250] is not preemptedby the FDCA or by a 2006 preamble to revised federalprescription drug regulations containing preemptivelanguage.

A. Statutory Preemption

*****OMITTED*****

[*62] C. The PLA's Presumption of Adequacy

The PLA, enacted in 1987, 33 codified liability on thepart of a manufacturer for failure to provide adequatewarnings, N.J.S.A. 2A:58C-2, and defined an adequateproduct warning as "one that a reasonably prudent personin the same or similar circumstances would haveprovided with respect to the danger and thatcommunicates adequate information on the dangers andsafe use of the product, . . . in the case of prescriptiondrugs, taking into account the characteristics of, and theordinary knowledge common to, the prescribingphysician." N.J.S.A. 2A:58C-4. That latter provisionadditionally establishes a rebuttable presumption that awarning, approved or prescribed by the FDA under theFDCA, is adequate. Ibid.

33 The PLA was not applicable in Feldman,which was initiated long before the Act's passage.

This presumption was construed, prior to Merck'swithdrawal of Vioxx, in Perez v. Wyeth Lab., 161 N.J. 1,734 A.2d 1245 (1999), a case alleging failure to directlywarn consumers of the difficulty of removing implants ofthe contraceptive Norplant. When reversing summaryjudgment in Wyeth's favor, the Court recognized a dutyto warn in direct-to-consumer [***79] advertising ofpharmaceuticals, but held that the presumption set forth

Page 17401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***in N.J.S.A. 2A:58C-4 was applicable in this context aswell. Id. at 21-25, 734 A.2d 1245. The Court found thatin this consumer context, "the same rebuttablepresumption should apply when a manufacturer complieswith FDA advertising, labeling and warningrequirements." Id. at 24, 734 A.2d 1245. The courtcontinued:

That approach harmonizes themanufacturer's duty to doctors and to thepublic when it chooses to directlyadvertise its products, and simultaneouslyrecognizes the public interest in informingpat ien ts [**256] about newpharmaceutical developments. Moreover,a rebuttable presumption that the duty toconsumers is met by compliance withFDA regulations helps to ensure thatmanufacturers are not made guarantorsagainst remotely possible, but notscientifically-verifiable, side-effects ofprescription drugs, a result that could havea "significant anti-utilitarian effect."

[Id. at 24-25, 734 A.2d 1245]

[*63] In language upon which defendant Merckstrongly relies, the Court then stated:

We believe that this standard is fair andbalanced. For all practical purposes,absent deliberate concealment ornondisclosure of af ter-acquiredknowledge of harmful effects, compliance [***80] with FDA standards should bevirtually dispositive of such [failure towarn] claims. By definition, theadvertising will have been "fairly'balanced."

[Id. at 25, 734 A.2d 1245]

See also Rowe v. Hoffman-La Roche, Inc., 189 N.J. 615,626, 917 A.2d 767 (2007) (utilizing this language fromPerez in a case concerning choice-of-law).

Merck claims on appeal that the language of Perezlimiting exceptions to the rebuttable presumption ofadequacy set forth in the PLA to instances of deliberateconcealment or nondisclosure, precludes liability in thiscase, because the results of its studies, particularly,VIGOR, were provided by Merck in a timely fashion tothe FDA and constituted a basis for the FDA's approvalof the revised Vioxx label. However, we are unwilling toconstrue the presumption as Merck urges, finding therecord in this case to be sufficient to support the

recognition of an additional basis for overcoming thepresumption of adequacy set forth in the PLA, applicableto Merck in the post-market warning context presentedhere. Specifically, we do not rest our decision torecognize this compensatory damage claim as one of"those rare cases when the presumption [of warningadequacy] is overcome," Perez, supra, 161 N.J. at 25,734 A.2d 1245 [***81] upon any claim of fraud on theFDA, thereby implicating the punitive damage aspects ofthe PLA. 34 Our focus rests solely upon plaintiffs' claimsof Merck's economically-driven manipulation of thepost-market regulatory process.

34 In this regard, we note that the Court in Perezrecognized that there could be circumstances inwhich a compensatory damage award wasappropriate, because the presumption of warningadequacy was overcome, but that a basis forpunitive damages would not exist. Ibid.

In concluding that a hitherto unrecognized legalbasis for an award of compensatory damages under thePLA exists here, we [*64] note that close scrutiny of theFDA and its regulatory power in a labeling contextcommenced only after Perez was decided, and thatscrutiny disclosed flaws in the regulatory system,existing at least until the time of the 2007 Amendments,35 that render the dictum of Perez less all-encompassingthan it might then have appeared. Commentators andcourts have since recognized that, whereas pre-marketapprovals of drugs are generally thorough in nature, theability of the FDA, post-market, "to detect unforeseenadverse effects of [a] drug and to take prompt andeffective remedial [***82] action" is considerably less.Kessler & Vladeck, supra, 96 Geo. L.J. at 465. It is theseflaws in that post-marketing oversight process thatprovide the foundation for the further exception to[**257] the presumption of adequacy that we findapplicable to this case. Kessler and Vladeck have stated:"Recent regulatory failures, such as the agency'sineffectual response to Vioxx, have demonstrated theFDA's shortcomings in this regard." Ibid. See alsoThomas N. Tiedt, The Drug Safety System Conundrum,62 Food & Drug L.J. 547, 551-55 (2007) (summarizingcriticisms of the FDA's post-market oversight). Thus,Kessler and Vladeck have asserted that on the day of newdrug approval, "and that day only, we agree that theFDA's determinations about labeling ought not be subjectto re-examination by courts or juries in failure-to-warncases." 96 Geo. L.J. at 465. Although, in light of thePLA's statutory presumption, we do not take so extremea position, we regard the scientific and regulatoryconditions upon which the authors then focus to behighly relevant to our consideration of whether the juryin this case could, on the basis of the evidence presentedand applicable law, determine that the presumption [***83] of adequacy had been overcome.

Page 18401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***

35 We express no opinion whether thestrengthening of the FDA's powers in 2007 willbe adequate to alleviate earlier-detectedproblems.

In this regard, Kessler and Vladeck first observe:

At the time of approval, the FDA'sknowledge-base may be close to perfect,but it is also highly limited because, atthat point, the drug has been tested on arelatively small population of patients.Once the drug enters the marketplace,risks that are [*65] relatively rare, thatmanifest themselves only after anextended period of time, or that affectvulnerable subpopulations, begin toemerge. These are often not risks foreseenby the drug's manufacturer or the FDAand, for that reason, are not addressed onthe label.

[Id. at 466 (footnotes omitted).]

See also U.S. Gov't Accountability Office, Drug Safety:Improvement Needed in FDA's Postmarket Decision-Making and Oversight Process, GAO 06-402 (2006)(hereafter, GAO Report) at 26 (discussing weaknesses inclinical trials); Comm. on the Assessment of the U.S.Drug Safety Sys., Inst. of Med. of the Nat'l Acads., TheFuture of Drug Safety: Promoting and Protecting theHealth of the Public, 37-39, 153 (Alina Baciu, KathleenStratton [***84] & Sheila P. Burke eds., 2006)(hereafter IOM Report); Tiedt, supra, 62 Food & DrugL.J. at 553. As the IOM Report's authors found: "It isworth underscoring that the fundamental design of thedrug approval system . . . -- separate from the quality ofthe data that sponsors generate in compliance with it --inevitably puts drugs on the market when safetyinformation is incomplete." IOM Report, supra, at 59.

Further, until the 2007 Amendments were passed,the FDA "did not have the [statutory] authority to compellabeling changes, but instead had to negotiate changeswith the drug's sponsor." Kessler & Vladeck, supra, 96Geo. L.J. at 466. As Kessler and Vladeck note inopposing preemption:

Manufacturers often resist labelingchanges the FDA believes are needed dueto emerging safety concerns. For instance,the FDA acknowledges that it took over ayear to force Merck, the manufacturer ofVioxx, to add a warning of the risks ofheart attack and stroke to Vioxx's label.During the lengthy negotiations, no

change was made to Vioxx's label, and inthe end, the FDA settled for a weakerwarning than it had proposed. As noted, atthe time of the Vioxx controversy, theFDA did not have statutory [***85]authority to compel manufacturers tomake labeling changes, but instead had torely on its power of persuasion, backed upby the FDA's authority to seek withdrawalof the drug's NDA or to file a misbrandingaction. The FDA generally got its way,but [**258] negotiations withmanufacturers are often quite lengthy andfrequently result in compromise decisions,as was the case with Vioxx.

[Id. at 480 (footnotes omitted).]

The FDA's Deputy Director of its Office of NewDrugs, Dr. Sandra Kweder, testified in a Senate hearingheld after the [*66] withdrawal of Vioxx that safetyconcerns over the drug prompted the FDA to convene anadvisory committee meeting in 2001 to determinewhether it increased the risk of heart attacks and strokes.Although the panel advised a change in the label toreflect that risk, the change was delayed. Additionally,Dr. Kweder acknowledged the lack of regulatoryauthority recognized by Kessler and Vladeck, stating:

[W]e don't have the authority to tell acompany, ["T]his is how your label has tolook. This is the language that needs to gointo your label. Here is where it goes, endof story.["] We have to negotiate with thecompany the specific language of howthings should be [***86] worded, theplacement, those kinds of things . . . .

* * *

[In connection with Vioxx, Merck]rejected many of our proposals, and wesimilarly rejected many of the proposals --most of the proposals they sent us.

[FDA's Drug Approval Process: Upto the Challenge?: Hearing Before the S.Comm. on Health, Educ. Labor andPensions, 109th Cong. 10, 26-27 (2005)(hereafter, Up to the Challenge).]

See also GAO Report, supra, at 10; IOM Report, supra,at 157-58. Dr. Kweder also acknowledged that the FDAlacked the power to compel additional post-marketingrandomized clinical trials or epidemiological studies.

We don't have the authority to tell them,

Page 19401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***you must do this particular trial. That is anauthority we don't have.

Now, we certainly have a fair amountof influence in convincing them to dosome of these studies, and we are, for themost part, reasonably successful. But wedon't have the authority to say, you mustdo the trial.

[Up to the Challenge, supra, at 23.]

See also GAO Report, supra, at 11, 27-28; IOM Report,supra, at 155-57.

Given these admitted flaws in the FDA's controlover postmarket labeling in the years that Vioxx was onthe market, we are unwilling to accept Merck's positionthat the [***87] presumption of adequacy of aprescription drug's label can be overcome only uponproof of deliberate concealment or nondisclosure. Factsunavailable to the Supreme Court at the time of the Perezdecision demonstrate that such a restriction is too narrow.

[*67] The FDCA requires federal approval of newdrugs, and mandates that, in order to obtain FDAapproval, a manufacturer must demonstrate thatadequate, well-controlled studies have demonstrated thedrug to be both safe and effective. 21 U.S.C.A. § 355.The "Indications and Usage" section of Merck's initiallabel stated that Vioxx was indicated for relief of thesigns and symptoms of osteoarthritis, for themanagement of acute pain in adults and for the treatmentof primary dysmenorrhea. Neither the "Warnings"section nor the "Precautions" section mentioned anyadverse cardiovascular effects.

At trial, plaintiffs took the position, supported bysufficient evidence, that by 1997, as the result ofFitzGerald's 023 study, Merck knew that Vioxxsuppressed prostacyclin and thus upset the balancebetween clotting and anti-clotting agents in the body. Italso knew that FitzGerald had postulated that theimbalance could lead to [**259] an increased risk [***88] of thrombotic events. 36 Studies at the time ofnew drug approval in 1999 arguably were insufficient toverify whether an increased cardiovascular risk existedfor patients taking Vioxx. But even then, the FDA'smedical review officer, Dr. Villalba, recognized anumerical increase in ischemic/thromboembolic events,and she recommended further cardiovascular testing.Such cardiovascular testing was also urged by Dr.Scolnick. However, despite preparatory steps, it was notconducted. Instead, the company focused on clinicaltrials intended to expand the market for its product,including the VIGOR trials.

36 We regard it to be immaterial whether the

FitzGerald hypothesis was correct, finding greatersignificance in the patent evidence of increasedCV risk from use of Vioxx -- whatever its cause.We note that the lack of specific CV studies byMerck has likely contributed to the absence ofspecific knowledge of causative factors.

Merck's VIGOR study confirmed the existence ofthe feared elevated thrombotic risk, as acknowledged bycompany officials in e-mails. Although Merck reportedthe VIGOR study results to the FDA on June 29, 2000 assupport for its supplemental new [*68] drug applicationand [***89] supplemented its report on October 13,2000, Merck's focus was on Vioxx's gastrointestinalsafety as compared with nonselective NSAIDs. At thattime, Merck sought to explain the adverse cardiovasculareffects disclosed by the study as consistent with the"known" anti-platelet effects of naproxen. However, nosuch effects, particularly effects of the magnituderequired to explain the difference in cardiovascularincidents, had been scientifically validated.

Further, although the FDA determined in February2001 that the results of the VIGOR study should beincorporated into the label for Vioxx and that a warningregarding cardiovascular risks should be expressed, analmost two-year period elapsed between the time thatMerck submitted its supplemental new drug application,intended to tout the GI benefits of Vioxx over traditionalNSAIDs, and the approval of the new label in April2002. The time span is even longer when calculated fromMarch 27, 2000, the date that Merck initially reported theresults of the VIGOR study to the FDA.

The record provides evidence sufficient to concludethat, during this period of time, Merck actively, and to anextent successfully, sought to dilute the labeling [***90]required as a result of the VIGOR study. Moreover,during this time, Merck's marketing personnel engaged instrenuous efforts to ensure that the results of the VIGORstudy were not communicated to prescribing physiciansby sales persons, and there is some evidentiary supportfor a claim of misrepresentation by Merck in respondingto individual physician inquiries. Additionally, althoughthe VIGOR results were published during this period, theincreased CV risk evident upon examination of eventsoccurring just after the study's CV cut-off date was notdisclosed in the published article. Further, the increasedrisk was not described as such, but rather framed in termsof the decreased incidence of cardiovascular thromboticevents associated with naproxen -- a traditional NSAIDimbued with cardioprotective powers whose extent, todate, remains unproven.

[*69] The fact that the label was finally revised inApril 2002 to reflect VIGOR's results, known to Merckat least by March 9, 2000 when Dr. Scolnickacknowledged the that "the CV events are clearly there,"provides powerful evidence that the label approved in

Page 20401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***May 1999, which contained no precautions or warningsregarding cardiovascular [**260] risks, [***91] wasinadequate, at least from March 9, 2000 onward.

We additionally find the evidence at trial sufficientto have permitted a jury to conclude that plaintiffs hadovercome the presumption of adequacy relating to therevised label approved in April 2002. In this regard, weparticularly note evidence of Merck's strenuous,economically driven, opposition to the inclusion ofcardiovascular risk in the "Warnings" section of theVioxx label, despite the universal opinions of the FDA'sadvisory committee and medical reviewers -- and indeed,initially, the FDA regulators, themselves -- that awarning was appropriate. That a lesser "Precaution,"limited only to patients with a history of ischemic, orpatent, heart disease, was approved can best be attributedto the dominant power of drug companies in a regulatoryprocess that permitted, and indeed required, efforts toresolve scientific disputes through conciliatory processes.

IV.

At the conclusion of the trial, the trial judgeinstructed the jury at length regarding Merck's duty towarn, incorporating in her instructions the PLA'srebuttable presumption of adequacy and applicablefederal labeling regulations, by stating:

You have heard a lot about [***92] theFDA's role. Under the Product LiabilityAct of New Jersey, which sets forth thelaw for failure to warn claims, there is aprovision that states that in the case of aclaim for failure to warn involving aprescription drug that there's a rebuttablepresumption that a label approved by theFDA is adequate. Therefore, we start withthe presumption that if the FDA approveda drug label, then the warnings in the labelare adequate.

However, if plaintiffs producesubstantial evidence that the [approved]label is not an adequate warning, then thepresumption can be overcome.

If plaintiffs produce such evidence,then you, the jury, must weigh all theevidence produced by both plaintiffs andthe defendant on the issue of the [*70] adequacy of the warning and decide ifplaintiffs have met [their] burden ofproving that Merck failed to provide anadequate warning to physicians. Thispresumption applies only to the label andonly where the FDA has approved thelabel as adequate.

However, if you find that the

plaintiffs have proven by a preponderanceof the evidence that after a label wasapproved there was new information thatchanged the known or knowablecardiovascular risks of VIOXX, thenunder [***93] FDA regulations, Merckhad a duty to warn physicians of anynewly discovered risks of the drug.

The FDA requires a drugmanufacturer to warn the medicalcommunity as soon as there's reasonableevidence of an association of a serioushazard with a drug, and that languagecomes from the FDA requirements andregulations.

There need not be proof of causation,only association. In other words, if there'sreasonable evidence of associationbetween taking a drug and certain harmoccurring without proof of exactly howthe drug causes the harm, the FDA stillrequires the warning be given to thephysicians of the risk.

Merck could, if it chooses to, withoutprior FDA approval send letters tophysicians, take out ads, publish injournals, or send out sales representativesin order to advise physicians of a newlyknown risk of VIOXX. There is aprocedure under the regulations, also,where a manufacturer of a drug likeMerck can change their label to add riskinformation and submit [it to] the FDA[**261] for approval within 30 days. Ifthe FDA doesn't object to the change inthat time, the new warning can be used.

* * *

It is up to you to decide what Merckknew or should have known aboutwhether there were [***94] potentialcardiovascular risks of VIOXX basedupon the reasonable evidence and when. Itis up to you to then determine whether inlight of all the information that Merckknew or should have known, it actedreasonably and adequately warnedphysicians of any serious cardiovascularrisks that they should have been warnedabout based on all the facts that you findto be true in the time period where theycould have gotten the information to theprescribing physician before the plaintiffs'heart attacks.

Page 21401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***

On appeal, Merck reiterates objections to thisinstruction that it made at trial, claiming that the chargedoes not reflect the Court's holding in Perez; assertingthat the judge erred in permitting the jury to consider thetimeliness of the 2002 label; and arguing further that thejudge's instructions "seriously misconstrued the FDAregulations" and, as the Court found in Feldman v.Lederle Labs. (Feldman III), 132 N.J. 339, 346-47, 625A.2d 1066 (1993), essentially directed a verdict againstMerck on the issue of its breach of a duty to warn.

[*71] We do not accept Merck's arguments. Theinstruction at issue adequately informed the jury that thepresumption of adequacy could only be overcome by"substantial [***95] evidence," thereby according thepresumption a significance greater than would otherwisebe the case, while not according it conclusive effect. SeePerez, supra, 161 N.J. at 24, 734 A.2d 1245 (citingFeldman II, supra, 125 N.J. at 156-57, 592 A.2d 1176);compare Shim v. Rutgers, 191 N.J. 374, 386, 924 A.2d465 (2007) ("[A] presumption has the effect ofcompelling a particular conclusion in the absence ofcontrary evidence. To overcome a presumption, evidencethat 'tends to' disprove the presumed fact, thereby raisinga debatable question regarding the existence of thepresumed fact, must be adduced."); N.J.R.E. 301.Although the instruction did not contain languagerestricting rebutting evidence to that relating to"deliberate concealment or nondisclosure," as Merckrequested, for the reasons that we have previouslyexplained, we do not accept that restriction as applicablein the present case.

Merck argues that the instruction improperlyintroduced the factor of the "timeliness" of the 2002 labelinto the case. But, in light of the labeling requirements of21 C.F.R. § 314.70(c)(2)(i) (permitting labeling changes"[t]o add or strengthen a contraindication, warning,precaution, or adverse reaction") and 21 C.F.R. §201.57(e) (specifying [***96] that a label "shall berevised to include a warning as soon as there isreasonable evidence of an association of a serious hazardwith a drug"), that issue was, properly, a principal focusof plaintiffs' proofs. See Feldman II, supra, 125 N.J. at157, 592 A.2d 1176 (noting that the rebuttablepresumption of the PLA "was enacted in the context ofpresent FDCA, PHSA [Public Health Service Act], andregulatory provisions that explicitly require warning ofpossible adverse side effects as soon as reasonablyfeasible and based on 'reasonable evidence.'"). Federalregulations, as well as the holding of Feldman II, wereaccurately described by the trial judge in an instructionthat focused both on the nature of the scientific evidencethat would trigger a duty to warn and the means by whichsuch a warning could be conveyed.

[*72] [**262] We also reject Merck's argument that

the instruction was fatally akin to that in Feldman III.There, the trial court instructed the jury that: "TheFederal Food and Drug Administration regulations andrequirements are minimal standards and the defendantstill owes a duty to warn its users in the exercise ofreasonable care." Feldman v. Lederle Lab., 257 N.J.Super. 163, 168, 608 A.2d 356 (App.Div.1992). [***97]The Supreme Court, agreeing with our analysis of theissue, determined that "[t]he trial court's error lay intelling the jury outright that Lederle had a 'duty to warn.'"Feldman III, supra, 132 N.J. at 347, 625 A.2d 1066. Suchlanguage does not appear in the charge given in this case,which properly placed upon plaintiffs the burden ofestablishing Merck's failure to provide an adequatewarning and appropriately directed the jury to considerwhat Merck knew or should have known, when factssufficient to require a warning became known, andwhether it acted reasonably, given the information that itpossessed.

V.

Merck additionally raises a number of evidentiaryarguments that we review under an abuse of discretionstandard.

*****OMITTED****VI.

At the conclusion of the evidence, the trial judgedirected a verdict for plaintiffs on the issue of whetherDr. Braun would have determined not to prescribe Vioxxto McDarby if adequately warned of its cardiovascularrisks (product-defect causation), recognizing theapplicability of a heeding presumption in thispharmaceutical context and determining that thepresumption had not been overcome. The judgeinstructed the jury:

If you find that Merck failed to providean adequate warning, then the lawrequires you to presume that plaintiffs'doctors would have heeded that adequatewarning and not have prescribed VIOXXto [plaintiffs]. However, to recoverdamages for their heart attacks, [plaintiffs]must still prove that their taking VIOXXwas [***112] a proximate cause of theirheart attacks.

In its new trial motion and on appeal, Merck arguesthat the heeding presumption is inapplicable topharmaceuticals and, if applicable, it was overcome.Accordingly, the trial judge erred.

Merck's arguments regarding the adoption of aheeding presumption in a pharmaceutical failure-to-warncontext essentially mirror those rejected by Judge Walsh

Page 22401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***when presented by Wyeth, Inc., in phen-fen litigationpending before him. See In re Diet Drug Litig., 384 N.J.Super. 525, 895 A.2d 480 (Law Div.2005). We agree inprinciple with Judge Walsh that in appropriatecircumstances, 40 a heeding presumption may beapplicable to claims of failure to warn of the dangers ofpharmaceuticals, as well as other products. In doing so,we find no basis to conclude that the Court's reasoning inCoffman v. Keene Corp., 133 N.J. 581, 595-603, 628A.2d 710 (1993), and Theer v. Philip Carey Co., 133 N.J.610, 618-24, 628 A.2d 724 (1993), should necessarily beinapplicable to a claim of failure to warn of the dangersof a [*81] palliative drug for which potentially lessharmful alternatives exist. 41 As the Court stated inCoffman:

The heeding presumption . . . serves toreinforce the basic duty to warn-toencourage manufacturers [***113] toproduce safer products, and to alert usersof the hazards arising from the use ofthose products through effective warnings.The duty to warn exists not only to protectand alert product users but to encouragemanufacturers and industries, whichbenefit from placing products into thestream of commerce, to remain apprisedof the hazards posed by a product. Theuse of the heeding presumption provides apowerful incentive for manufacturers toabide by their duty to provide adequatewarnings. See Nissen Trampoline Co. v.Terre Haute First Nat'l Bank, 332 N.E.2d820, 826 (Ind. Ct. App. 1975) (holdingthat heeding presumption [**268] "woulddiscourage those manufacturers whowould rather risk liability than provide awarning which would impair themarketability of the product"), rev'd onprocedural grounds,[265 Ind. 457] 358N.E.2d 974 (1976).

[133 N.J. at 599, 628 A.2d 710.]

That comment is equally apt in a pharmaceutical context.

40 In light of the analysis that follows, we donot find it necessary to establish in this opinionwhat such circumstances would be. 41 Although, as Merck argues, thecardiovascular risk to McDarby from continueduse of Vioxx was "unavoidable," the use of adifferent palliative agent provided an alternative [***114] means for pain relief.

We attribute no particular significance to the factthat the heeding presumption was not mentioned by the

Court in Strumph v. Schering Corp., 133 N.J. 33, 626A.2d 1090 (1993), reversing for the reasons expressed byJudge Skillman in his dissent, 256 N.J. Super. 309, 323,606 A.2d 1140 (App.Div.1992), a prescription drug casealleging failure to warn that was decided in the sameterm as Coffman and Theer. In light of testimony inStrumph by both treating physicians that they were awareof the risks of the drug that they prescribed and, havingconducted a risk-benefit analysis, nonetheless determinedits use to be warranted, Strumph, supra, 256 N.J. Super.at 323-24, 606 A.2d 1140, use of such a presumptionwould not have been factually sustainable or, analyzedotherwise, the presumption would have been rebutted asa matter of law.

Merck argues additionally that, because of the risk-benefit analysis that physicians undertake whenprescribing medications, "one cannot 'presume' thatadditional risk information would lead a prescribingphysician to avoid the drug." Recognition of that [*82] circumstance is incorporated into the generally rebuttablenature of the heeding presumption, permitting a drugmanufacturer [***115] to counter a plaintiff's causationargument with contrary evidence, as in fact occurred inStrumph. Thus, the heeding presumption does not stifleinnovation, as Merck suggests, but merely fosters thedisclosure of accurate information regarding risk on new,as well as established, pharmaceutical products.

However, we do agree with Merck that, inMcDarby's case, the judge's use of the heedingpresumption in her legal analysis and jury instructionswas not legally required. That presumption, precedentdemonstrates, is primarily applicable in circumstances inwhich plaintiff lacks the ability to prove by directevidence that a proper warning, if given, would havebeen heeded. Coffman, supra, 133 N.J. at 600, 628 A.2d710. But here, direct evidence in the form of thedeposition testimony of McDarby's treating physicianexisted, rendering use of a presumption unnecessary.Nonetheless, we do not regard the judge's use ofpresumption language to have resulted in reversibleerror, since we are satisfied that directing a verdict onthis causation issue was proper. As the Court has held,"in the absence of any countervailing evidence, 'a trialjudge need not submit the issue of proximate cause fromthe absence [***116] of a warning to the jury but maydetermine as a matter of law that the warning would havebeen heeded.'" Coffman, supra, 133 N.J. at 595, 628 A.2d710 (quoting Coffman v. Keene Corp., 257 N.J. Super.279, 290, 608 A.2d 416 (App.Div.1992)). That isessentially what the trial judge did here in directing averdict in plaintiffs' favor on this causation issue.

Our review of the record satisfies us that the judgeruled appropriately in this regard. Dr. Braun's depositiontestimony discloses his close attention to Merck's productliterature, including its package inserts, "Dear Doctor"

Page 23401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***letters, and the CV card, and his reliance upon Merck'sassurances of safety in the face of the published results ofthe VIGOR trial and the questions regarding thecardiovascular risks of Vioxx posed by Dr. Topol. Thedoctor's testimony also demonstrates that, when informedby [*83] Merck that [**269] Vioxx posed a risk topatients with ischemic heart disease, the doctordiscontinued prescribing the drug to a patient with thatcondition. He thus followed Merck's instructions whereapplicable, demonstrating his willingness to cease the useof a popular and effective medication, but on the basis ofhis treatment records, determined the inapplicability [***117] of Merck's precautions to McDarby. As a finalmatter, Dr. Braun testified unequivocally that he wouldnot have added to the cardiovascular risks confrontingMcDarby as the result of his age, gender and diabeticcondition if he had known Vioxx "could" increase therisk of a heart attack. Similarly, McDarby testified thathe would not have taken the drug if he had known of itscardiovascular risk, and he stated that he relied on hisdoctor for a determination of drug safety.

Merck argues, nonetheless, that Dr. Braun was neverasked whether he would have ceased prescribing Vioxxto McDarby if adequately warned of an "association"between Vioxx and an increased risk of seriouscardiovascular events. While we recognize the scientificdistinction between a causal relationship and anassociative one, we do not regard this linguistic quibbleas sufficient to have raised a jury issue, given thestrength of Dr. Braun's testimony in this case.Additionally, Merck argues that a jury could have foundthat, after April 2002, Dr. Braun would have continued toprescribe a drug that had proven effective, noting thatMcDarby needed pain relief, he had taken the drugwithout problems for two years, he [***118] was takingcardioprotective aspirin, and debates still existedregarding the cardiovascular safety of Vioxx. However,this argument is wholly speculative, and finds no supportin the unequivocal testimony given by Dr. Braun.

VII.

At trial, testimony was presented by plaintiffs'experts, Dr. Krumholz, and cardiologist Dr. NicholasDePace 42 to establish [*84] that Vioxx was a substantialcontributing factor in the heart attack suffered byMcDarby on April 14, 2004. Dr. Krumholz testified, inaccordance with the FitzGerald hypothesis, that Vioxx'saction as a COX-2 inhibitor was thought 43 to upset thebody's balance between prostacyclin and thromboxane byinhibiting prostacyclin production, thereby increasing theclotting action of platelets in the blood that would occurwhen plaque deposited in arteries ruptured, and that theincreased clotting could lead to blockage of the normalblood flow and the occurrence of a heart attack. Dr.Krumholz testified further that the risk that such clottingwould lead to a heart attack was increased in patients

with other elevated risk factors such as atherosclerosis,elevated "bad" cholesterol levels, or diabetes, utilizing agraphic illustration from a [***119] 2005 article in thejournal Circulation 44 to demonstrate the impact of COX-2 inhibition on clotting in patients with atheroscleroticblood vessels.

42 Dr. DePace is board-certified in cardiology.He serves as a clinical professor at the ThomasJefferson Medical School in Philadelphia and aphysician at the Jefferson Heart Center.43 The doctor recognized the existence of otherhypotheses, but found this was supported by the"most evidence," had "gotten the most attention"and was the one that the scientific communitywas "most concerned about." 44 Elliott M. Antman, David DeMets & JosephLoscalzo, Cyclooxygenase Inhibition andCardiovascular Risk, 112 Circulation, 759(2005).

In additional testimony, Dr. Krumholz described theresults of the APPROVe study, which disclosed a relativerisk of adverse thrombotic cardiovascular events fromuse of Vioxx of 1.92. The doctor [**270] testified thatthe risk would be further elevated in diabetics, stating:

It is reasonably probable that diabeticsare at greater risk from VIOXX becausethey have an underlying higher risk ofdisease. Diabetes is a risk factor for heartdisease . . . . VIOXX . . . would be moredangerous in that group in absolute termsthan [***120] it would be in the othergroup.

Whereas studies had shown an elevated risk of heartdisease among diabetics of 1.5, it was Dr. Krumholz'sopinion that a "conservative estimate" would place theincreased risk to a diabetic taking Vioxx at "at least twotimes the risk." Although the doctor testified that theprecipitating cause of a heart attack [*85] (whether age,diabetes, low "good" cholesterol, or Vioxx) could not bephysically identified, the existing scientific studies haddemonstrated that a forty-eight-month history of use ofVioxx would constitute a substantial contributing factorto its occurrence.

Dr. DePace, who had examined McDarby,confirmed the presence of risk factors in addition to long-term use of Vioxx, consisting of his age, low levels of"good" cholesterol, weight, and diabetes, and he alsoconcluded that Vioxx had been a substantial contributingfactor to his heart attack. In reaching this conclusion, thedoctor relied upon the existing epidemiological studies,including VIGOR and APPROVe. In addition to thegeneral results of the APPROVe study, indicating a 1.92

Page 24401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***relative risk of a serious thrombotic event, the doctornoted that the authors of the APPROVe study hadconducted [***121] a subgroup analysis of patientstaking Vioxx who had a history of diabetes that discloseda relative risk of 6.10 or a 510% increase in risk.Although the doctor recognized that the post hocsubgroup analysis had limitations, he nonetheless foundthe findings to be significant.

Merck's expert, Dr. Barry Rayburn, conceded oncross-examination that McDarby had taken Vioxx forforty-eight months before his heart attack and, whereasthe APPROVe study indicated an overall relative risk of1.92 for serious thrombotic events, a post hoc analysisshowed an elevation of that relative risk to 4.45 inpatients taking the drug for nineteen to thirty-six months,for a 345% increase in risk.

On appeal, Merck contends that evidence of anincrease in relative risk such as that to which the expertstestified was insufficient to establish causation. Merckparticularly challenges any reliance on the subgroupanalysis performed by its scientists on the APPROVedata. However, that evidence did not constitute the solebasis for the opinions of either of plaintiffs' experts, andthe potential lack of reliability of the subgroup analysiswas exhaustively demonstrated to the jury. Ampleevidence supported an [***122] increased risk resultingfrom the conjoined effects of diabetes and Vioxx,whether the jury accepted the more conservative [*86] estimates of Dr. Krumholz or the higher estimates thatDr. DePace considered in reaching his opinion. Thatepidemiological evidence, combined with theexplanatory opinions of both Dr. Krumholz and Dr.Deface were sufficient to create the jury issue regardingcausation. Landrigan, supra, 127 N.J. at 412-23, 605A.2d 1079; see also Grassis v. Johns-Manville Corp.,248 N.J. Super. 446, 454-56, 591 A.2d 671(App.Div.1991).

We reject Merck's argument, premised on theCourt's decision in Cruz-Mendez v. ISU/Ins. Servs. of SanFrancisco, 156 N.J. 556, 722 A.2d 515 (1999), that thejury had to find "but for" causation and that the judgeerred in not giving that instruction. In Cruz-Mendez, acase involving the misuse by plaintiff of fireworks foundafter a display, an issue existed whether the fireworks[**271] display was the proximate cause of theplaintiff's injury or whether plaintiff's conduct afterfinding the fireworks constituted an intervening cause sounforeseeable that the causal chain was broken. Id. at576, 722 A.2d 515. In this circumstance, the Courtreversed a determination that as a matter of law, plaintiffhad demonstrated [***123] causation "because thefirework that injured his hand 'was attributable to afireworks display that was put on approximately fivedays earlier.'" Id. at 574, 722 A.2d 515. The Court heldthat plaintiff must show both that "defendant's act or

omission was the factual, or 'but for,' cause of the injury"and that this factual cause was a proximate cause of theinjury. Ibid.

However, as the Court explained in Verdicchio v.Ricca, 179 N.J. 1, 843 A.2d 1042 (2004):

[T]he "but for" test has its limitations insituations where two or more forcesoperate to bring about a certain result and"any one of them operating alone wouldbe sufficient." Indeed, the "but for" testhas been characterized as a potentially"insurmountable obstacle" for a plaintiffin a case in which "unrelated factors mayhave contributed to the same injury."

In response to the apparent limitationof the "but for" test in concurrentcausation cases, New Jersey, like manyjurisdictions, has adopted a modifiedstandard -- the substantial factor standard-- "limited to that class of cases in which adefendant's negligence combines with apreexistent condition to cause harm -- asdistinguished from cases in which thedeviation alone is the cause of the harm."

[Id. at 24, 843 A.2d 1042 [***124](citations omitted).]

[*87] Thus, the language of Cruz-Mendez isinapplicable in a case such as this in which multiplefactors could be found by a jury to have contributed toMcDarby's condition. In this matter, medical causationwas appropriately demonstrated by proof that exposure tothe defendant's product "was a substantial factor incausing or exacerbating the disease." James v. BessemerProcessing Co., 155 N.J. 279, 299, 714 A.2d 898 (1998)(quoting Sholtis v. Am. Cyanamid Co., 238 N.J. Super. 8,30-31, 568 A.2d 1196 (App.Div.1989)) (adoptingstandard in toxic tort context); 45 see also Model JuryCharge (Civil), 612, "Proximate Cause -- Where There isClaim that Concurrent Causes of Harm Were Present"(1998). In sum, in this case there was adequate proof ofMcDarby's continued, long-term use of Vioxx and"medical and/or scientific proof of a nexus between [thatuse] and . . . plaintiff's condition." James, supra, 155 N.J.at 304, 714 A.2d 898. Thus, the jury could properlyconclude, as it did, that medical causation had beendemonstrated. We thus affirm the compensatory damageaward by the jury in connection with McDarby's cause ofaction for failure to warn in violation of the PLA.

45 To the extent that the requirement in James [***125] of proof of frequency, regularity andproximity, id. at 302-04, 714 A.2d 898, is

Page 25401 N.J. Super. 10, *; 949 A.2d 223, **;

2008 N.J. Super. LEXIS 116, ***imported into this drug context, we find thatstandard met by the uncontroverted proof of useby McDarby of Vioxx for a period of forty-eightmonths.

VIII.

The PLA provides:

Punitive damages shall not be awardedif a drug or device . . . *******OMITTED*****

In summary, we affirm the award ofcompensatory damages to McDarby pursuant to

the PLA, determining that the cause of actionasserted under that statute is not preempted andthat no [*99] reversible error occurred inconnection with that claim. We reverse the awardof punitive damages pursuant to the PLA aspreempted by the FDCA, and we reverse theawards of damages to McDarby and Cona and theawards of attorneys' fees pursuant to the CFA,determining that plaintiffs' CFA claims aresubsumed within the PLA.

Affirmed in part and reversed in part.