The discussion about the CV effects of coxibs really started after the publication of the VIGOR study. It may be worth while reminding oneself that this was really a GI safety study, and of course not a CV study - neither by design or actual patient years. VIGOR demonstrated VIOXX superior GI safety on all pre specified endpoints The bar chart here are depicting the events rate per 100 patient year for three of the key endpoints with the relative risk and 95%CI of the relative risk above the bars. For all three key endpoints, confirmed clinical upper GI events, confirmed Complicated events and all episodes of clinical GI bleeding there were highly significant reductions of 54, 57 and 62% respectively.
In VIGOR, there were 45 confirmed thrombotic events on rofecoxib and 19 on naproxen. Therefore the relative risk of sustaining a confirmed CV event on naproxen compared with rofecoxib was 0.42 with 95% CI which do not cross 1 which implies statistical significance. Although there was a reduction in confirmed CV events, there was no difference in CV mortality. Seven patients died from a cardiovascular event in each group. If you break these events down by location you can see that the majority of events were cardiac events. The relative risk of sustaining a cardiac event on naproxen compared with rofecoxib was 0.36. Cardiac events drove the analyses. Within the cardiac event category, most of the events were myocardial infarctions and there was a significant reduction in myocardial infarctions on naproxen compared to rofecoxib. To better understand these results we looked at the clinical characteristics of patients with events. We found that the patients who had thrombotic events were those who you would have expected to have events--they were older, there was a higher percentage of males, and close to 80% had one or more CV risk factors
On the vertical axis is the cumulative incidence of investigator reported CV events with time on the x-axis. In green is the combined NSAID group from the Phase IIB/III studies and orange is the rofecoxib group. Overlaid on top in yellow if the rofecoxib group from VIGOR with the blue showing the VIGOR naproxen group. What you see if that the rates of events in the rofecoxib group from VIGOR and Phase IIb/III and the combined NSADI group are virtually superimposable, the outlier is the naproxen group which appears to have a lower incidence of events compared to the other three groups.