Evidence Based Assessment of the Cardiovascular Adverse Effects of Specific Cyclooxy-Genase-2 Inhibitors

Abstract

Objective : To provide a quality assessment and a systematic review of all the meta analysis and systematic reviews of the cardiovascular and related safety of rofecoxib for its licensed long-term indications. Method : Systematically search Pubmed, Cochrane, FDA, NICE, EULAR, CCOHTA, to identify systematic reviews concerning rofecoxib’s cardiovascular toxicity. The resulting set of citations was hand searched by title, MESH, and whenever available abstract. Reference lists of recovered articles were further screened for any additional citations. Quality was assessed using the QUOROM (Quality of Reporting of Meta analyses) checklist. Quality scores were divided into quartiles (<25% i.e. very poor,25-49% i.e. poor, 50-75% i.e. acceptable and >75% i.e. good quality) to assist judgement of quality. Results : 15 systematic reviews were included with a total of 43,343 patients. The mean overall quality of reporting score is acceptable [63.13%, (SD: 21.41%, range =33.3-94.4%)]. The overall score for Title (27.30%) and Results (45.46%) were poor, while overall scores for Abstracts (69.70%), Methods (71.21%) and Discussion(63.60%) were acceptable. Nine systematic reviews utilised manufacturer’s files and 8of these either received funding or the authors were employed by the manufacturer. Only 4 systematic reviews reported an end search date after 2000, when the results from the VIGOR trial became available. Cumulative meta-analysis was not performed prior to rofecoxib’s withdrawal. Rofecoxib was compared with few active comparators whose cardiovascular safety is not established. Majority of patients were female, Caucasians and younger than in actual practice (<65 years old) and exposed to rofecoxib for a median of only 6 weeks, a duration inadequate to assess long-termsafety. Conclusions : The worldwide withdrawal of rofecoxib emphasises further the need for adherence to standardised reporting and quality guidelines to allow researchers to synthesise available information in quantitative and unbiased manner allowing for timely and appropriate decisions. Limitation of available evidence combined with alack of adequate pharmacovigilace independent from manufacturer’s interest delayed the recognition of rofecoxib’s cardiovascular adverse events. Design of prospective cardiovascular toxicity evaluations should test patients that resemble the population likely to receive the drug in clinical practice using composite outcomes for at least a year comparing the agent with placebo and established comparators.