Observational Studies Published after the Withdrawal of Rofecoxib

OBSERVATIONAL STUDIES PUBLISHED AFTER THE WITHDRAWAL OF ROFECOXIB

Eleven observational studies on COX-2 inhibitors and adverse cardiovascular outcomes were reported after the withdrawal of rofecoxib and major findings are summarized in Table 47.2. Reports in abstract form or conference proceedings are not included in this review. The results were heterogeneous as there was much variation in study design, study populations, comparison groups, and outcomes of interest. Four reports were based on administrative data of private or public health insurance data in North America. Shaya and colleagues compared incidence of APTC events among users of COX-2 inhibitors and non-selective NSAIDs and found no increased risk among users of rofecoxib or celecoxib when compared with users of non-selective NSAIDs (Shaya et al., 2005).

Graham and colleagues evaluated the risk of myocar-dial infarction and sudden cardiac death among users of non-selective NSAIDs and COX-2 inhibitors in a nested case-control study (Graham et al., 2005). Comparing with those who were unlikely to have used a prescription non-selective NSAID or COX-2 inhibitor during the last 60 days, the adjusted odds ratio was 0.84 (95% CI, 0.67–1.04) for celecoxib. Adjusted odds ratio was higher for high dose rofe-coxib (more than 25 mg per day) than for low dose rofecoxib (25 mg per day or less). Levesque and colleagues compared current use of COX-2 inhibitors, non-selective NSAIDs, and no use of either COX-2 inhibitors or non-selective NSAIDs and found that high dose rofecoxib (more than 25 mg per day) and low dose rofecoxib (25 mg per day or less) were both associated with increased risk of hospitalized myocar-dial infarction when compared with no use (Levesque, Brophy and Zhang, 2005). There was no increased risk among celecoxib users. Velentgas and colleagues reported an increased risk of acute coronary syndrome among current rofecoxib users compared with ibupro-fen or diclofenac users (relative risk 1.35; 95% CI, 1.09–1.68) and no increased risk among current cele-coxib users for the same comparison drugs (relative risk 1.03; 95% CI, 0.83–1.27) (Velentgas et al., 2006). Huang and colleagues used national health insurance data from Taiwan and reported no increased risk of myocardial infarction, angina, stroke, or transient ischemic attack when rofecoxib or celecoxib was indi-vidually compared with meloxicam among an ethnic Chinese population (Huang et al., 2006).

Two studies were based on electronic medical record systems in the UK. Hippisley-Cox and Coup-land found that rofecoxib use was associated with myocardial infarction (adjusted odds ratio 1.32; 95% CI, 1.09–1.61) and celecoxib use showed similar level of increased risk, but the lower bound of the 95% CI was 0.97 (Hippisley-Cox and Coupland, 2005). Andersohn studied stroke as an outcome interest and found increased risk of stroke among rofecoxib and etoricoxib users but not celecoxib users (Andersohn et al., 2006).

Two studies were based on population-based registries in Denmark. Johnsen and colleagues stud-ied the risk of first myocardial infarction and reported increased risk among rofecoxib users and non-statistically- significant increased risk among celecoxib users (Johnsen et al., 2005). Gislason and colleagues studied re-infarction and death after post-myocardial infarction discharge and found increased risk for both celecoxib and rofecoxib (Gislason et al., 2006).

The two other studies were case–control studies with patients identified from hospitals. Kimmel and colleagues in the US compared the use of COX-2 inhibitors with no NSAID use and reported adjusted odds ratio of 0.43 (95% CI, 0.23–0.79) for celecoxib and 1.16 (95% CI, 0.70–1.93) for rofecoxib (Kimmel et al., 2005). McGettigan and colleagues in Australia studied acute coronary syndrome and did not find increased risk among celecoxib users or rofecoxib users (McGettigan, Han and Henry, 2006).

McGettigan and Henry combined data from 12 observational studies involving COX-2 inhibitors and found that celecoxib was not associated with increased cardiovascular risk, with a combined rela-tive risk of 1.06 (95% CI, 0.91–1.23) (McGettigan and Henry, 2006). Both high dose rofecoxib (more than 25 mg per day) and lower dose rofecoxib (25 mg or less per day) were associated with increased cardiovascular risk. Combined relative risk for high rofecoxib was 2.19 (95% CI, 1.64–2.91) and it was 1.33 (95% CI, 1.00–1.79) for lower dose rofecoxib.