New England Journal of Medicine Exposé on Drug Scandals

The Journal played a centralrole in two prominent drug-safe-ty developments of the 21st cen-tury, involving the cyclooxygen-ase-2 inhibitor rofecoxib (Vioxx)and the oral hypoglycemic agentrosiglitazone (Avandia). Accord-

stroke (2005a). Accumulating evi-dence clarified that the VIGORfindings on myocardial infarctionwere not attributable to a cardio-protective effect of naproxen but toa cardiotoxic effect of rofecoxib.Follow-up publications drew atten-tion to other important problemswith the depiction of adverseevents in that trial, including theselective omission of key adverseevents caused by the sponsor’sdrug and an earlier cutoff date

for recording side effectsthat cast the product in anunfavorable light (myocardialinfarctions) than for thoserevealing its clinical advan-tage (gastrointestinal bleedingevents). This selective report-ing prompted the editors towrite two “Expressions ofConcern” about the report-ing of risk data in that trial(2005b, 2006).

Recent years have seenthe growth of a new mecha-nism for the study of ad-verse drug events: court ac-

tion forcing the release of rawclinical trial data held by thecompanies that funded the stud-ies.2 Years after publication of theoriginal Vioxx studies, such re-analyses of their underlying trialreports have found clear evidence,from 3 years before the drug’swithdrawal, of an increased riskof cardiovascular death,3 as wellas evidence that its gastroprotec-tive advantage for most patientswas greatly overstated.4 The ad-vent of mandatory adverse-eventreporting at ClinicalTrials.gov islikely in the coming years to pro-vide an additional means of en-suring more timely detection ofdrug-safety problems (2011a).

Rofecoxib became anothermilestone in the punctuated evo-lution of drug-safety science and

ing to the pivotal study promot-ing rofecoxib, the Vioxx Gastro-intestinal Outcomes Research(VIGOR) trial (2000), patientsgiven that drug had significantlyfewer episodes of gastrointesti-nal bleeding than those takingnaproxen. The authors also re-ported an incidental finding:naproxen users had one fourth asmany myocardial infarctions aspatients in the rofecoxib group.They speculated that the causewas a cardioprotective antiplate-let effect of naproxen.

The VIGOR article becamenotorious, for several reasons. In2004, another randomized trialof rofecoxib was stopped earlybecause the drug, as comparedwith placebo, nearly doubled therisk of myocardial infarction and

An Early 20th-Century Ad Featuring Drugs fromBayer, Including Heroin for Coughs and Aspirin.